You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by tq...@apache.org on 2022/06/01 14:36:19 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@0cd4dd2f2d6cab265844de0cb8745e0de8d22571)
This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/tvm-site.git
The following commit(s) were added to refs/heads/asf-site by this push:
new ceb1466ae deploying docs (apache/tvm@0cd4dd2f2d6cab265844de0cb8745e0de8d22571)
ceb1466ae is described below
commit ceb1466ae21437d837493eddbb0b6d32516e4865
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Wed Jun 1 14:36:06 2022 +0000
deploying docs (apache/tvm@0cd4dd2f2d6cab265844de0cb8745e0de8d22571)
---
.../how_to/compile_models/from_mxnet.rst.txt | 2 +-
.../how_to/compile_models/from_oneflow.rst.txt | 2 +-
.../how_to/compile_models/from_paddle.rst.txt | 2 +-
.../how_to/compile_models/from_pytorch.rst.txt | 2 +-
.../compile_models/sg_execution_times.rst.txt | 22 +-
.../deploy_models/deploy_model_on_android.rst.txt | 2 +-
.../deploy_object_detection_pytorch.rst.txt | 4 +-
.../deploy_models/deploy_prequantized.rst.txt | 6 +-
.../deploy_prequantized_tflite.rst.txt | 4 +-
.../how_to/deploy_models/deploy_quantized.rst.txt | 2 +-
.../deploy_models/deploy_ssd_gluoncv.rst.txt | 4 +-
.../deploy_models/sg_execution_times.rst.txt | 18 +-
.../extend_tvm/bring_your_own_datatypes.rst.txt | 2 +-
.../how_to/extend_tvm/sg_execution_times.rst.txt | 10 +-
.../how_to/extend_tvm/use_pass_instrument.rst.txt | 16 +-
.../optimize_operators/opt_conv_cuda.rst.txt | 2 +-
.../optimize_operators/opt_conv_tensorcore.rst.txt | 2 +-
.../how_to/optimize_operators/opt_gemm.rst.txt | 16 +-
.../optimize_operators/sg_execution_times.rst.txt | 8 +-
.../sg_execution_times.rst.txt | 16 +-
.../tune_conv2d_layer_cuda.rst.txt | 1071 +++++++++--
.../tune_network_cuda.rst.txt | 2 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 124 +-
.../tune_with_autotvm/sg_execution_times.rst.txt | 12 +-
.../tune_with_autotvm/tune_conv2d_cuda.rst.txt | 34 +-
.../work_with_microtvm/micro_autotune.rst.txt | 16 +-
.../work_with_microtvm/sg_execution_times.rst.txt | 10 +-
.../work_with_relay/sg_execution_times.rst.txt | 8 +-
.../work_with_schedules/sg_execution_times.rst.txt | 18 +-
.../how_to/work_with_schedules/tensorize.rst.txt | 2 +-
.../tutorials/autotvm/sg_execution_times.rst.txt | 6 +-
.../frontend/deploy_classification.rst.txt | 2 +-
.../tutorials/frontend/deploy_detection.rst.txt | 2 +-
.../tutorials/frontend/sg_execution_times.rst.txt | 6 +-
.../tutorials/optimize/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/sg_execution_times.rst.txt | 6 +-
.../tutorial/auto_scheduler_matmul_x86.rst.txt | 2 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 54 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 24 +-
.../tutorial/tensor_expr_get_started.rst.txt | 47 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 127 +-
docs/how_to/compile_models/from_paddle.html | 2 +-
docs/how_to/compile_models/from_pytorch.html | 4 +-
docs/how_to/compile_models/sg_execution_times.html | 22 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 17 +-
docs/how_to/deploy_models/deploy_prequantized.html | 6 +-
.../deploy_models/deploy_prequantized_tflite.html | 4 +-
docs/how_to/deploy_models/deploy_quantized.html | 2 +-
docs/how_to/deploy_models/deploy_ssd_gluoncv.html | 34 +-
docs/how_to/deploy_models/sg_execution_times.html | 18 +-
.../extend_tvm/bring_your_own_datatypes.html | 2 +-
docs/how_to/extend_tvm/sg_execution_times.html | 10 +-
docs/how_to/extend_tvm/use_pass_instrument.html | 16 +-
docs/how_to/optimize_operators/opt_conv_cuda.html | 2 +-
.../optimize_operators/opt_conv_tensorcore.html | 2 +-
docs/how_to/optimize_operators/opt_gemm.html | 16 +-
.../optimize_operators/sg_execution_times.html | 8 +-
.../sg_execution_times.html | 14 +-
.../tune_conv2d_layer_cuda.html | 1071 +++++++++--
.../tune_with_autoscheduler/tune_network_cuda.html | 2 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 124 +-
.../tune_with_autotvm/sg_execution_times.html | 12 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.html | 34 +-
docs/how_to/work_with_microtvm/micro_autotune.html | 16 +-
.../work_with_microtvm/sg_execution_times.html | 10 +-
.../how_to/work_with_relay/sg_execution_times.html | 8 +-
.../work_with_schedules/sg_execution_times.html | 18 +-
docs/how_to/work_with_schedules/tensorize.html | 2 +-
docs/reference/api/doxygen/affine__type_8h.html | 2 +-
.../api/doxygen/affine__type_8h__incl.svg | 1014 ++++++-----
docs/reference/api/doxygen/algorithm_8h__incl.svg | 520 +++---
docs/reference/api/doxygen/algorithms_8h__incl.svg | 308 ++--
.../api/doxygen/analyzer_8h__dep__incl.svg | 644 +++----
docs/reference/api/doxygen/analyzer_8h__incl.svg | 204 +--
docs/reference/api/doxygen/annotation_8h.html | 2 +-
docs/reference/api/doxygen/annotation_8h__incl.svg | 1066 +++++------
.../api/doxygen/apply__history__best_8h__incl.svg | 528 +++---
.../api/doxygen/arg__info_8h__dep__incl.svg | 108 +-
docs/reference/api/doxygen/arg__info_8h__incl.svg | 340 ++--
.../api/doxygen/array__utils_8h__dep__incl.svg | 44 +-
.../api/doxygen/array__utils_8h__incl.svg | 432 ++---
.../api/doxygen/auto__schedule_8h__incl.svg | 384 ++--
.../auto__scheduler_2cost__model_8h__incl.svg | 272 +--
.../doxygen/auto__scheduler_2feature_8h__incl.svg | 532 +++---
docs/reference/api/doxygen/autodiff_8h__incl.svg | 284 +--
docs/reference/api/doxygen/bias__add_8h__incl.svg | 684 +++----
docs/reference/api/doxygen/bitserial_8h__incl.svg | 212 +--
.../api/doxygen/block__scope_8h__dep__incl.svg | 128 +-
.../api/doxygen/block__scope_8h__incl.svg | 134 +-
docs/reference/api/doxygen/bound_8h__dep__incl.svg | 628 +++----
docs/reference/api/doxygen/bound_8h__incl.svg | 220 +--
.../api/doxygen/broadcast_8h__dep__incl.svg | 92 +-
docs/reference/api/doxygen/broadcast_8h__incl.svg | 708 ++++----
docs/reference/api/doxygen/buffer_8h.html | 2 +-
.../reference/api/doxygen/buffer_8h__dep__incl.svg | 640 +++----
docs/reference/api/doxygen/buffer_8h__incl.svg | 1080 +++++------
.../api/doxygen/builder_8h__dep__incl.svg | 40 +-
docs/reference/api/doxygen/builder_8h__incl.svg | 368 ++--
.../api/doxygen/builtin_8h__dep__incl.svg | 108 +-
docs/reference/api/doxygen/builtin_8h__incl.svg | 332 ++--
docs/reference/api/doxygen/call_8h.html | 2 +-
docs/reference/api/doxygen/call_8h__incl.svg | 1066 +++++------
docs/reference/api/doxygen/codegen_8h__incl.svg | 416 ++---
docs/reference/api/doxygen/codegen_8h_source.html | 2 +-
.../doxygen/compilation__config_8h__dep__incl.svg | 20 +-
.../api/doxygen/compilation__config_8h__incl.svg | 344 ++--
.../api/doxygen/compute__dag_8h__dep__incl.svg | 76 +-
.../api/doxygen/compute__dag_8h__incl.svg | 476 ++---
.../api/doxygen/constant__utils_8h__dep__incl.svg | 168 +-
.../api/doxygen/constant__utils_8h__incl.svg | 756 ++++----
.../reference/api/doxygen/cublas_8h__dep__incl.svg | 20 +-
docs/reference/api/doxygen/cublas_8h__incl.svg | 468 ++---
.../api/doxygen/cuda_2dense_8h__dep__incl.svg | 12 +-
.../reference/api/doxygen/cuda_2dense_8h__incl.svg | 724 ++++----
.../api/doxygen/cuda_2injective_8h__dep__incl.svg | 12 +-
.../api/doxygen/cuda_2injective_8h__incl.svg | 748 ++++----
.../api/doxygen/cuda_2pooling_8h__dep__incl.svg | 12 +-
.../api/doxygen/cuda_2pooling_8h__incl.svg | 720 ++++----
.../api/doxygen/cuda_2reduction_8h__dep__incl.svg | 12 +-
.../api/doxygen/cuda_2reduction_8h__incl.svg | 748 ++++----
.../api/doxygen/cuda_2softmax_8h__dep__incl.svg | 12 +-
.../api/doxygen/cuda_2softmax_8h__incl.svg | 748 ++++----
.../api/doxygen/data__layout_8h__dep__incl.svg | 112 +-
.../api/doxygen/data__layout_8h__incl.svg | 328 ++--
.../api/doxygen/database_8h__dep__incl.svg | 20 +-
docs/reference/api/doxygen/database_8h__incl.svg | 552 +++---
.../api/doxygen/dataflow__matcher_8h__incl.svg | 608 +++----
.../doxygen/dataflow__pattern_8h__dep__incl.svg | 24 +-
.../api/doxygen/dataflow__pattern_8h__incl.svg | 640 +++----
.../dataflow__pattern__functor_8h__dep__incl.svg | 12 +-
.../dataflow__pattern__functor_8h__incl.svg | 616 +++----
.../doxygen/detail_2broadcast_8h__dep__incl.svg | 104 +-
.../api/doxygen/detail_2broadcast_8h__incl.svg | 588 +++---
.../api/doxygen/detail_2extern_8h__dep__incl.svg | 40 +-
.../api/doxygen/detail_2extern_8h__incl.svg | 540 +++---
docs/reference/api/doxygen/device__api_8h.html | 2 +-
.../api/doxygen/device__api_8h__dep__incl.svg | 64 +-
.../reference/api/doxygen/device__api_8h__incl.svg | 622 +++----
.../api/doxygen/device__copy_8h__incl.svg | 312 ++--
.../api/doxygen/diagnostic_8h__dep__incl.svg | 536 +++---
docs/reference/api/doxygen/diagnostic_8h__incl.svg | 260 +--
docs/reference/api/doxygen/dilate_8h__incl.svg | 468 ++---
.../reference/api/doxygen/driver__api_8h__incl.svg | 588 +++---
docs/reference/api/doxygen/einsum_8h__incl.svg | 684 +++----
.../api/doxygen/elemwise_8h__dep__incl.svg | 48 +-
docs/reference/api/doxygen/elemwise_8h__incl.svg | 320 ++--
docs/reference/api/doxygen/env__func_8h.html | 2 +-
.../api/doxygen/env__func_8h__dep__incl.svg | 556 +++---
docs/reference/api/doxygen/env__func_8h__incl.svg | 738 ++++----
docs/reference/api/doxygen/error_8h__dep__incl.svg | 524 +++---
docs/reference/api/doxygen/error_8h__incl.svg | 260 +--
docs/reference/api/doxygen/executable_8h.html | 2 +-
.../api/doxygen/executable_8h__dep__incl.svg | 12 +-
docs/reference/api/doxygen/executable_8h__incl.svg | 1002 +++++-----
.../api/doxygen/executable_8h_source.html | 4 +-
docs/reference/api/doxygen/executor_8h__incl.svg | 332 ++--
.../api/doxygen/extracted__task_8h__incl.svg | 412 ++---
.../api/doxygen/feature__extractor_8h__incl.svg | 484 ++---
docs/reference/api/doxygen/flatten_8h__incl.svg | 592 +++---
docs/reference/api/doxygen/functions_func_t.html | 4 +-
docs/reference/api/doxygen/functions_func_v.html | 4 +-
docs/reference/api/doxygen/functions_s.html | 6 +-
docs/reference/api/doxygen/functions_t.html | 4 +-
docs/reference/api/doxygen/functions_v.html | 12 +-
docs/reference/api/doxygen/fuse_8h__dep__incl.svg | 156 +-
docs/reference/api/doxygen/fuse_8h__incl.svg | 432 ++---
.../api/doxygen/generic_2default_8h__incl.svg | 748 ++++----
.../api/doxygen/generic_2extern_8h__dep__incl.svg | 24 +-
.../api/doxygen/generic_2extern_8h__incl.svg | 736 ++++----
.../doxygen/generic_2injective_8h__dep__incl.svg | 32 +-
.../api/doxygen/generic_2injective_8h__incl.svg | 748 ++++----
docs/reference/api/doxygen/generic__func_8h.html | 2 +-
.../api/doxygen/generic__func_8h__dep__incl.svg | 196 +-
.../api/doxygen/generic__func_8h__incl.svg | 1912 ++++++++++----------
docs/reference/api/doxygen/greedy_8h__incl.svg | 420 ++---
docs/reference/api/doxygen/image_8h__incl.svg | 212 +--
docs/reference/api/doxygen/index__map_8h.html | 2 +-
.../api/doxygen/index__map_8h__dep__incl.svg | 568 +++---
docs/reference/api/doxygen/index__map_8h__incl.svg | 1082 +++++------
docs/reference/api/doxygen/instruction_8h.html | 2 +-
.../api/doxygen/instruction_8h__dep__incl.svg | 148 +-
.../reference/api/doxygen/instruction_8h__incl.svg | 740 ++++----
docs/reference/api/doxygen/instrument_8h.html | 2 +-
.../api/doxygen/instrument_8h__dep__incl.svg | 516 +++---
docs/reference/api/doxygen/instrument_8h__incl.svg | 750 ++++----
.../api/doxygen/int__set_8h__dep__incl.svg | 660 +++----
docs/reference/api/doxygen/int__set_8h__incl.svg | 190 +-
.../reference/api/doxygen/int__solver_8h__incl.svg | 412 ++---
.../reference/api/doxygen/interpreter_8h__incl.svg | 468 ++---
docs/reference/api/doxygen/ir_2adt_8h.html | 2 +-
.../api/doxygen/ir_2adt_8h__dep__incl.svg | 620 +++----
docs/reference/api/doxygen/ir_2adt_8h__incl.svg | 1088 +++++------
docs/reference/api/doxygen/ir_2attrs_8h.html | 2 +-
.../api/doxygen/ir_2attrs_8h__dep__incl.svg | 580 +++---
docs/reference/api/doxygen/ir_2attrs_8h__incl.svg | 1060 +++++------
docs/reference/api/doxygen/ir_2expr_8h.html | 2 +-
.../api/doxygen/ir_2expr_8h__dep__incl.svg | 604 +++----
docs/reference/api/doxygen/ir_2expr_8h__incl.svg | 994 +++++-----
docs/reference/api/doxygen/ir_2function_8h.html | 2 +-
.../api/doxygen/ir_2function_8h__dep__incl.svg | 628 +++----
.../api/doxygen/ir_2function_8h__incl.svg | 1100 +++++------
.../api/doxygen/ir_2module_8h__dep__incl.svg | 616 +++----
docs/reference/api/doxygen/ir_2module_8h__incl.svg | 260 +--
.../reference/api/doxygen/ir_2op_8h__dep__incl.svg | 576 +++---
docs/reference/api/doxygen/ir_2op_8h__incl.svg | 332 ++--
docs/reference/api/doxygen/ir_2span_8h.html | 2 +-
.../api/doxygen/ir_2span_8h__dep__incl.svg | 708 ++++----
docs/reference/api/doxygen/ir_2span_8h__incl.svg | 856 ++++-----
.../api/doxygen/ir_2transform_8h__dep__incl.svg | 536 +++---
.../api/doxygen/ir_2transform_8h__incl.svg | 312 ++--
docs/reference/api/doxygen/ir_2type_8h.html | 2 +-
.../api/doxygen/ir_2type_8h__dep__incl.svg | 696 +++----
docs/reference/api/doxygen/ir_2type_8h__incl.svg | 894 ++++-----
.../api/doxygen/iter__affine__map_8h__incl.svg | 416 ++---
.../api/doxygen/libtorch__runtime_8h__incl.svg | 604 ++++---
.../api/doxygen/libtorch__runtime_8h_source.html | 2 +-
.../api/doxygen/local__response__norm_8h__incl.svg | 444 ++---
.../api/doxygen/loop__state_8h__dep__incl.svg | 88 +-
.../reference/api/doxygen/loop__state_8h__incl.svg | 332 ++--
docs/reference/api/doxygen/mapping_8h__incl.svg | 444 ++---
.../api/doxygen/measure_8h__dep__incl.svg | 48 +-
docs/reference/api/doxygen/measure_8h__incl.svg | 328 ++--
.../doxygen/measure__callback_8h__dep__incl.svg | 12 +-
.../api/doxygen/measure__callback_8h__incl.svg | 516 +++---
.../api/doxygen/measure__record_8h__incl.svg | 332 ++--
docs/reference/api/doxygen/memory__pools_8h.html | 2 +-
.../api/doxygen/memory__pools_8h__dep__incl.svg | 44 +-
.../api/doxygen/memory__pools_8h__incl.svg | 1855 ++++++++++---------
.../meta__schedule_2cost__model_8h__dep__incl.svg | 12 +-
.../meta__schedule_2cost__model_8h__incl.svg | 484 ++---
docs/reference/api/doxygen/metadata_8h__incl.svg | 120 +-
docs/reference/api/doxygen/metadata__base_8h.html | 2 +-
.../api/doxygen/metadata__base_8h__dep__incl.svg | 12 +-
.../api/doxygen/metadata__base_8h__incl.svg | 1034 +++++------
.../api/doxygen/mutator_8h__dep__incl.svg | 32 +-
docs/reference/api/doxygen/mutator_8h__incl.svg | 412 ++---
docs/reference/api/doxygen/nn_2bnn_8h__incl.svg | 632 +++----
.../api/doxygen/nn_2dense_8h__dep__incl.svg | 24 +-
docs/reference/api/doxygen/nn_2dense_8h__incl.svg | 444 ++---
.../reference/api/doxygen/nn_2pooling_8h__incl.svg | 760 ++++----
.../reference/api/doxygen/nn_2softmax_8h__incl.svg | 744 ++++----
docs/reference/api/doxygen/node_8h.html | 2 +-
docs/reference/api/doxygen/node_8h__dep__incl.svg | 676 +++----
docs/reference/api/doxygen/node_8h__incl.svg | 836 ++++-----
docs/reference/api/doxygen/on__device_8h__incl.svg | 312 ++--
.../api/doxygen/op__strategy_8h__incl.svg | 772 ++++----
.../api/doxygen/operation_8h__dep__incl.svg | 592 +++---
docs/reference/api/doxygen/operation_8h__incl.svg | 432 ++---
docs/reference/api/doxygen/packed__func_8h.html | 2 +-
.../api/doxygen/packed__func_8h__dep__incl.svg | 544 +++---
.../api/doxygen/packed__func_8h__incl.svg | 726 ++++----
.../api/doxygen/packed__func_8h_source.html | 4 +-
.../api/doxygen/pad__utils_8h__dep__incl.svg | 12 +-
docs/reference/api/doxygen/pad__utils_8h__incl.svg | 432 ++---
docs/reference/api/doxygen/papi_8h.html | 2 +-
docs/reference/api/doxygen/papi_8h__incl.svg | 772 ++++----
docs/reference/api/doxygen/parser_8h__incl.svg | 368 ++--
docs/reference/api/doxygen/pattern_8h__incl.svg | 190 +-
.../api/doxygen/pattern__functor_8h__incl.svg | 584 +++---
.../api/doxygen/postproc_8h__dep__incl.svg | 32 +-
docs/reference/api/doxygen/postproc_8h__incl.svg | 412 ++---
docs/reference/api/doxygen/profiling_8h.html | 2 +-
.../api/doxygen/profiling_8h__dep__incl.svg | 12 +-
docs/reference/api/doxygen/profiling_8h__incl.svg | 760 ++++----
.../reference/api/doxygen/profiling_8h_source.html | 2 +-
docs/reference/api/doxygen/random_8h.html | 2 +-
docs/reference/api/doxygen/random_8h__incl.svg | 1070 +++++------
.../api/doxygen/ravel__unravel_8h__dep__incl.svg | 84 +-
.../api/doxygen/ravel__unravel_8h__incl.svg | 432 ++---
docs/reference/api/doxygen/reduce_8h.html | 2 +-
docs/reference/api/doxygen/reduce_8h__incl.svg | 1066 +++++------
.../api/doxygen/reduction_8h__dep__incl.svg | 40 +-
docs/reference/api/doxygen/reduction_8h__incl.svg | 768 ++++----
docs/reference/api/doxygen/reflection_8h.html | 2 +-
.../api/doxygen/reflection_8h__dep__incl.svg | 700 +++----
docs/reference/api/doxygen/reflection_8h__incl.svg | 720 ++++----
docs/reference/api/doxygen/registry_8h.html | 2 +-
.../api/doxygen/registry_8h__dep__incl.svg | 580 +++---
docs/reference/api/doxygen/registry_8h__incl.svg | 758 ++++----
.../api/doxygen/relay_2adt_8h__dep__incl.svg | 36 +-
docs/reference/api/doxygen/relay_2adt_8h__incl.svg | 480 ++---
.../api/doxygen/relay_2analysis_8h__incl.svg | 556 +++---
.../api/doxygen/relay_2attrs_2debug_8h.html | 2 +-
.../api/doxygen/relay_2attrs_2debug_8h__incl.svg | 1082 +++++------
.../api/doxygen/relay_2attrs_2memory_8h__incl.svg | 596 +++---
.../api/doxygen/relay_2attrs_2nn_8h__incl.svg | 212 +--
.../relay_2attrs_2transform_8h__dep__incl.svg | 20 +-
.../doxygen/relay_2attrs_2transform_8h__incl.svg | 520 +++---
.../reference/api/doxygen/relay_2attrs_2vm_8h.html | 2 +-
.../api/doxygen/relay_2attrs_2vm_8h__incl.svg | 1068 +++++------
.../api/doxygen/relay_2base_8h__dep__incl.svg | 296 +--
.../reference/api/doxygen/relay_2base_8h__incl.svg | 222 +--
.../api/doxygen/relay_2expr_8h__dep__incl.svg | 212 +--
.../reference/api/doxygen/relay_2expr_8h__incl.svg | 568 +++---
.../api/doxygen/relay_2expr__functor_8h__incl.svg | 584 +++---
.../api/doxygen/relay_2feature_8h__incl.svg | 444 ++---
.../api/doxygen/relay_2function_8h__dep__incl.svg | 36 +-
.../api/doxygen/relay_2function_8h__incl.svg | 520 +++---
.../api/doxygen/relay_2op_8h__dep__incl.svg | 36 +-
docs/reference/api/doxygen/relay_2op_8h__incl.svg | 636 +++----
.../relay_2op__attr__types_8h__dep__incl.svg | 28 +-
.../doxygen/relay_2op__attr__types_8h__incl.svg | 756 ++++----
.../api/doxygen/relay_2qnn_2attrs_8h.html | 2 +-
.../api/doxygen/relay_2qnn_2attrs_8h__incl.svg | 1066 +++++------
.../api/doxygen/relay_2qnn_2transform_8h__incl.svg | 632 +++----
.../api/doxygen/relay_2transform_8h__dep__incl.svg | 12 +-
.../api/doxygen/relay_2transform_8h__incl.svg | 636 +++----
.../api/doxygen/relay_2type_8h__dep__incl.svg | 240 +--
.../reference/api/doxygen/relay_2type_8h__incl.svg | 484 ++---
docs/reference/api/doxygen/reorg_8h__incl.svg | 812 ++++-----
.../api/doxygen/rocblas_8h__dep__incl.svg | 12 +-
docs/reference/api/doxygen/rocblas_8h__incl.svg | 468 ++---
.../reference/api/doxygen/rocm_2dense_8h__incl.svg | 728 ++++----
.../api/doxygen/rocm_2injective_8h__incl.svg | 748 ++++----
.../api/doxygen/rocm_2pooling_8h__incl.svg | 724 ++++----
.../api/doxygen/rocm_2reduction_8h__incl.svg | 748 ++++----
.../api/doxygen/rocm_2softmax_8h__incl.svg | 748 ++++----
.../reference/api/doxygen/runner_8h__dep__incl.svg | 76 +-
docs/reference/api/doxygen/runner_8h__incl.svg | 296 +--
docs/reference/api/doxygen/runtime_2module_8h.html | 3 +-
.../api/doxygen/runtime_2module_8h__dep__incl.svg | 544 +++---
.../api/doxygen/runtime_2module_8h__incl.svg | 632 +++----
.../api/doxygen/runtime_2module_8h_source.html | 44 +-
docs/reference/api/doxygen/runtime_2vm_2vm_8h.html | 2 +-
.../api/doxygen/runtime_2vm_2vm_8h__incl.svg | 1342 +++++++-------
.../api/doxygen/runtime_2vm_2vm_8h_source.html | 2 +-
docs/reference/api/doxygen/runtime_8h__incl.svg | 332 ++--
.../api/doxygen/schedule__pass_8h__dep__incl.svg | 136 +-
.../api/doxygen/schedule__pass_8h__incl.svg | 440 ++---
.../api/doxygen/schedule__rule_8h__dep__incl.svg | 32 +-
.../api/doxygen/schedule__rule_8h__incl.svg | 412 ++---
docs/reference/api/doxygen/search/all_11.js | 2 +-
docs/reference/api/doxygen/search/all_13.js | 2 +-
docs/reference/api/doxygen/search/all_14.js | 10 +-
docs/reference/api/doxygen/search/all_15.js | 4 +-
docs/reference/api/doxygen/search/all_16.js | 4 +-
docs/reference/api/doxygen/search/all_17.js | 4 +-
docs/reference/api/doxygen/search/all_4.js | 2 +-
docs/reference/api/doxygen/search/all_e.js | 4 +-
docs/reference/api/doxygen/search/functions_10.js | 2 +-
docs/reference/api/doxygen/search/functions_13.js | 2 +-
docs/reference/api/doxygen/search/functions_14.js | 4 +-
docs/reference/api/doxygen/search/functions_15.js | 2 +-
docs/reference/api/doxygen/search/functions_16.js | 2 +-
docs/reference/api/doxygen/search/functions_3.js | 2 +-
docs/reference/api/doxygen/search/functions_d.js | 2 +-
.../api/doxygen/search__policy_8h__dep__incl.svg | 12 +-
.../api/doxygen/search__policy_8h__incl.svg | 356 ++--
.../api/doxygen/search__strategy_8h__dep__incl.svg | 56 +-
.../api/doxygen/search__strategy_8h__incl.svg | 484 ++---
.../api/doxygen/search__task_8h__dep__incl.svg | 60 +-
.../api/doxygen/search__task_8h__incl.svg | 372 ++--
docs/reference/api/doxygen/source__map_8h.html | 2 +-
.../api/doxygen/source__map_8h__dep__incl.svg | 620 +++----
.../reference/api/doxygen/source__map_8h__incl.svg | 974 +++++-----
.../api/doxygen/space__generator_8h__dep__incl.svg | 32 +-
.../api/doxygen/space__generator_8h__incl.svg | 380 ++--
docs/reference/api/doxygen/state_8h__dep__incl.svg | 120 +-
docs/reference/api/doxygen/state_8h__incl.svg | 404 ++---
docs/reference/api/doxygen/stmt_8h__dep__incl.svg | 596 +++---
docs/reference/api/doxygen/stmt_8h__incl.svg | 134 +-
.../api/doxygen/stmt__functor_8h__dep__incl.svg | 12 +-
.../api/doxygen/stmt__functor_8h__incl.svg | 280 +--
.../api/doxygen/strided__slice_8h__dep__incl.svg | 72 +-
.../api/doxygen/strided__slice_8h__incl.svg | 396 ++--
docs/reference/api/doxygen/tag_8h__incl.svg | 340 ++--
.../reference/api/doxygen/target_8h__dep__incl.svg | 504 +++---
docs/reference/api/doxygen/target_8h__incl.svg | 420 ++---
docs/reference/api/doxygen/target__info_8h.html | 2 +-
.../api/doxygen/target__info_8h__incl.svg | 1010 ++++++-----
.../api/doxygen/target__kind_8h__dep__incl.svg | 504 +++---
.../api/doxygen/target__kind_8h__incl.svg | 236 +--
.../api/doxygen/task__scheduler_8h__incl.svg | 524 +++---
.../api/doxygen/te_2schedule_8h__dep__incl.svg | 648 +++----
.../api/doxygen/te_2schedule_8h__incl.svg | 328 ++--
.../reference/api/doxygen/tensor_8h__dep__incl.svg | 664 +++----
docs/reference/api/doxygen/tensor_8h__incl.svg | 388 ++--
.../api/doxygen/tensor__intrin_8h__dep__incl.svg | 640 +++----
.../api/doxygen/tensor__intrin_8h__incl.svg | 392 ++--
docs/reference/api/doxygen/tensor__type_8h.html | 2 +-
.../api/doxygen/tensor__type_8h__dep__incl.svg | 248 +--
.../api/doxygen/tensor__type_8h__incl.svg | 1014 ++++++-----
.../api/doxygen/tensor__utils_8h__dep__incl.svg | 80 +-
.../api/doxygen/tensor__utils_8h__incl.svg | 432 ++---
.../api/doxygen/tir_2analysis_8h__dep__incl.svg | 176 +-
.../api/doxygen/tir_2analysis_8h__incl.svg | 520 +++---
docs/reference/api/doxygen/tir_2expr_8h.html | 2 +-
.../api/doxygen/tir_2expr_8h__dep__incl.svg | 612 +++----
docs/reference/api/doxygen/tir_2expr_8h__incl.svg | 1210 +++++++------
.../doxygen/tir_2expr__functor_8h__dep__incl.svg | 20 +-
.../api/doxygen/tir_2expr__functor_8h__incl.svg | 134 +-
.../api/doxygen/tir_2function_8h__dep__incl.svg | 508 +++---
.../api/doxygen/tir_2function_8h__incl.svg | 196 +-
.../api/doxygen/tir_2op_8h__dep__incl.svg | 660 +++----
docs/reference/api/doxygen/tir_2op_8h__incl.svg | 456 ++---
.../api/doxygen/tir_2op__attr__types_8h.html | 2 +-
.../doxygen/tir_2op__attr__types_8h__dep__incl.svg | 184 +-
.../api/doxygen/tir_2op__attr__types_8h__incl.svg | 1016 ++++++-----
.../tir_2schedule_2schedule_8h__dep__incl.svg | 112 +-
.../doxygen/tir_2schedule_2schedule_8h__incl.svg | 412 ++---
.../api/doxygen/tir_2transform_8h__incl.svg | 392 ++--
.../api/doxygen/tir_2usmp_2analysis_8h__incl.svg | 404 ++---
.../api/doxygen/tir_2usmp_2transform_8h__incl.svg | 308 ++--
.../api/doxygen/tir_2usmp_2utils_8h__dep__incl.svg | 36 +-
.../api/doxygen/tir_2usmp_2utils_8h__incl.svg | 316 ++--
.../api/doxygen/topi_2nn_8h__dep__incl.svg | 12 +-
docs/reference/api/doxygen/topi_2nn_8h__incl.svg | 756 ++++----
.../api/doxygen/topi_2transform_8h__dep__incl.svg | 64 +-
.../api/doxygen/topi_2transform_8h__incl.svg | 728 ++++----
docs/reference/api/doxygen/topi_2utils_8h.html | 2 +-
.../reference/api/doxygen/topi_2utils_8h__incl.svg | 1008 ++++++-----
docs/reference/api/doxygen/trace_8h.html | 2 +-
docs/reference/api/doxygen/trace_8h__dep__incl.svg | 140 +-
docs/reference/api/doxygen/trace_8h__incl.svg | 748 ++++----
.../api/doxygen/transform__step_8h__dep__incl.svg | 96 +-
.../api/doxygen/transform__step_8h__incl.svg | 340 ++--
.../api/doxygen/tune__context_8h__dep__incl.svg | 24 +-
.../api/doxygen/tune__context_8h__incl.svg | 432 ++---
.../api/doxygen/type__functor_8h__incl.svg | 632 +++----
.../api/doxygen/type__relation_8h__dep__incl.svg | 564 +++---
.../api/doxygen/type__relation_8h__incl.svg | 324 ++--
docs/reference/api/doxygen/var_8h.html | 2 +-
docs/reference/api/doxygen/var_8h__dep__incl.svg | 616 +++----
docs/reference/api/doxygen/var_8h__incl.svg | 1020 +++++------
.../api/doxygen/virtual__device_8h__dep__incl.svg | 256 +--
.../api/doxygen/virtual__device_8h__incl.svg | 364 ++--
docs/reference/api/doxygen/vision_8h__incl.svg | 212 +--
docs/reference/api/doxygen/x86_2bnn_8h__incl.svg | 744 ++++----
.../api/doxygen/x86_2default_8h__incl.svg | 748 ++++----
.../api/doxygen/x86_2injective_8h__incl.svg | 744 ++++----
docs/reference/api/python/auto_scheduler.html | 4 +-
.../api/typedoc/classes/bytestreamreader.html | 12 +-
.../api/typedoc/classes/cachedcallstack.html | 34 +-
docs/reference/api/typedoc/classes/dldatatype.html | 12 +-
docs/reference/api/typedoc/classes/dldevice.html | 10 +-
.../reference/api/typedoc/classes/environment.html | 12 +-
docs/reference/api/typedoc/classes/ffilibrary.html | 20 +-
.../api/typedoc/classes/graphexecutor.html | 16 +-
docs/reference/api/typedoc/classes/instance.html | 40 +-
docs/reference/api/typedoc/classes/memory.html | 34 +-
docs/reference/api/typedoc/classes/module.html | 10 +-
docs/reference/api/typedoc/classes/ndarray.html | 22 +-
.../api/typedoc/classes/packedfunccell.html | 6 +-
docs/reference/api/typedoc/classes/rpcserver.html | 14 +-
docs/reference/api/typedoc/classes/scalar.html | 6 +-
.../api/typedoc/classes/webgpucontext.html | 12 +-
docs/reference/api/typedoc/enums/argtypecode.html | 30 +-
.../api/typedoc/enums/aynccallbackcode.html | 4 +-
.../api/typedoc/enums/dldatatypecode.html | 8 +-
.../api/typedoc/enums/rpcserverstate.html | 12 +-
docs/reference/api/typedoc/enums/sizeof.html | 18 +-
docs/reference/api/typedoc/index.html | 112 +-
.../api/typedoc/interfaces/disposable.html | 2 +-
.../api/typedoc/interfaces/functioninfo.html | 6 +-
.../api/typedoc/interfaces/libraryprovider.html | 4 +-
docs/searchindex.js | 2 +-
.../vta/tutorials/autotvm/sg_execution_times.html | 6 +-
.../tutorials/frontend/deploy_classification.html | 2 +-
.../vta/tutorials/frontend/deploy_detection.html | 2 +-
.../vta/tutorials/frontend/sg_execution_times.html | 6 +-
.../vta/tutorials/optimize/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/sg_execution_times.html | 6 +-
docs/tutorial/auto_scheduler_matmul_x86.html | 2 +-
docs/tutorial/autotvm_relay_x86.html | 110 +-
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 24 +-
docs/tutorial/tensor_expr_get_started.html | 43 +-
475 files changed, 70081 insertions(+), 68285 deletions(-)
diff --git a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
index 245c19091..550bf7030 100644
--- a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
@@ -98,7 +98,7 @@ In this section, we download a pretrained imagenet model and classify an image.
.. code-block:: none
- Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipe4724699-b020-4e8b-9f7b-3ca7e7462716 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip1f7a3cf1-7d31-4f66-a000-26fa89c1904f from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
x (1, 3, 224, 224)
diff --git a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
index 4f3cb3a67..706b3fcc6 100644
--- a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
@@ -100,7 +100,7 @@ Load a pretrained OneFlow model and save model
.. code-block:: none
Downloading: "https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip" to /workspace/.oneflow/flowvision_cache/resnet18.zip
-
0%| | 0.00/41.5M [00:00<?, ?B/s]
0%| | 16.0k/41.5M [00:00<07:34, 95.6kB/s]
0%| | 48.0k/41.5M [00:00<04:47, 151kB/s]
0%| | 104k/41.5M [00:00<03:05, 234kB/s]
0%| | 160k/41.5M [00:00<02:38, 273kB/s]
1%| | 224k/41.5M [00:00<02:18, 312kB/s]
1%| | 288k/41.5M [00:01<02:08, 335kB/s]
1%| | 360k/41.5M [00:01<01:58, 365kB/s]
1%|1 | 432k/41.5M [00:01<01:51, 385kB/s]
1%|1 | 512k/41.5M [00:01<01:43, 413kB/s]
1%|1 | 592k/41.5M [00:01<01:39, 433kB/s]
2%|1 | 680k/41.5M [00:01<01:33, 460kB/s]
2%|1 | 768k/41.5M [00:02<01:29, 479kB/s]
2%|2 | 864k/41.5M [00:02<01:24, 507kB/s]
2%|2 | 960k/41.5M [00:02<01:20, 526kB/s]
3%|2 | 1.04M/41.5M [00:02<01:16, 554kB/s]
3%|2 | 1.15M/41.5M [00:02<01:12, 587kB/s]
3%|3 | 1.26M/41.5M [00:02<01:09, 611kB/s]
3%|3 | 1.38M/41.5M [00:03<01:05, 642kB/s]
4%|3 | 1.49M/41.5M [00:03<01:03, 663kB/s]
4%|3 | 1.62M/41.5M [00:03<01:00, 692kB/s]
4%|4 | 1.75M/41.5M [00:03<00:57, 727kB/s]
5%|4 | 1.88M/41.5M [00:03<00:55, 751kB/s]
5%|4 | 2.02M/41.5M [00:03<00:52, 783kB/s]
5%|5 | 2.17M/41.5M [00:04<00:50, 819kB/s]
6%|5 | 2.33M/41.5M [00:04<00:47, 858kB/s]
6%|6 | 2.49M/41.5M [00:04<00:45, 900kB/s]
6%|6 | 2.66M/41.5M [00:04<00:43, 944kB/s]
7%|6 | 2.84M/41.5M [00:04<00:41, 988kB/s]
7%|7 | 3.03M/41.5M [00:04<00:39, 1.03MB/s]
8%|7 | 3.23M/41.5M [00:05<00:36, 1.09MB/s]
8%|8 | 3.44M/41.5M [00:05<00:35, 1.14MB/s]
9%|8 | 3.66M/41.5M [00:05<00:33, 1.19MB/s]
9%|9 | 3.88M/41.5M [00:05<00:31, 1.25MB/s]
10%|9 | 4.12M/41.5M [00:05<00:29, 1.32MB/s]
11%|# | 4.38M/41.5M [00:06<00:2
8, 1.38MB/s]
11%|#1 | 4.63M/41.5M [00:06<00:26, 1.44MB/s]
12%|#1 | 4.91M/41.5M [00:06<00:25, 1.50MB/s]
13%|#2 | 5.19M/41.5M [00:06<00:24, 1.57MB/s]
13%|#3 | 5.48M/41.5M [00:06<00:23, 1.64MB/s]
14%|#3 | 5.80M/41.5M [00:06<00:21, 1.72MB/s]
15%|#4 | 6.13M/41.5M [00:07<00:20, 1.81MB/s]
16%|#5 | 6.48M/41.5M [00:07<00:19, 1.90MB/s]
16%|#6 | 6.84M/41.5M [00:07<00:18, 1.98MB/s]
17%|#7 | 7.22M/41.5M [00:07<00:17, 2.09MB/s]
18%|#8 | 7.62M/41.5M [00:07<00:16, 2.19MB/s]
19%|#9 | 8.03M/41.5M [00:07<00:14, 2.49MB/s]
20%|## | 8.47M/41.5M [00:08<00:12, 2.75MB/s]
21%|##1 | 8.75M/41.5M [00:08<00:13, 2.62MB/s]
22%|##1 | 9.01M/41.5M [00:08<00:13, 2.46MB/s]
23%|##2 | 9.41M/41.5M [00:08<00:12, 2.71MB/s]
24%|##3 | 9.91M/41.5M [00:08<00:11, 2.82MB/s]
25%|##5 | 10.4M/41.5M [00:08<00:10, 3.20MB/s]
26%|##5
| 10.8M/41.5M [00:08<00:10, 3.07MB/s]
27%|##6 | 11.1M/41.5M [00:08<00:11, 2.86MB/s]
28%|##7 | 11.5M/41.5M [00:09<00:10, 2.90MB/s]
29%|##9 | 12.1M/41.5M [00:09<00:09, 3.11MB/s]
31%|### | 12.7M/41.5M [00:09<00:09, 3.29MB/s]
32%|###2 | 13.4M/41.5M [00:09<00:08, 3.46MB/s]
34%|###3 | 14.0M/41.5M [00:09<00:08, 3.58MB/s]
35%|###5 | 14.6M/41.5M [00:09<00:06, 4.04MB/s]
37%|###6 | 15.3M/41.5M [00:10<00:05, 4.58MB/s]
38%|###8 | 15.8M/41.5M [00:10<00:06, 4.20MB/s]
39%|###9 | 16.2M/41.5M [00:10<00:06, 3.94MB/s]
41%|#### | 16.8M/41.5M [00:10<00:06, 3.84MB/s]
42%|####2 | 17.6M/41.5M [00:10<00:06, 4.11MB/s]
44%|####4 | 18.4M/41.5M [00:10<00:05, 4.38MB/s]
46%|####6 | 19.2M/41.5M [00:10<00:04, 4.90MB/s]
48%|####8 | 20.1M/41.5M [00:11<00:04, 5.58MB/s]
50%|####9 | 20.6M/41.5M [00:11<00:04, 5.29MB/s]
51%|##### | 21.2M/41.5M [00:11<00:
04, 4.94MB/s]
53%|#####3 | 22.0M/41.5M [00:11<00:03, 5.37MB/s]
55%|#####5 | 23.0M/41.5M [00:11<00:03, 6.18MB/s]
57%|#####6 | 23.6M/41.5M [00:11<00:03, 6.09MB/s]
58%|#####8 | 24.2M/41.5M [00:11<00:03, 5.57MB/s]
61%|###### | 25.1M/41.5M [00:11<00:02, 6.61MB/s]
62%|######2 | 25.8M/41.5M [00:12<00:02, 6.50MB/s]
64%|######3 | 26.5M/41.5M [00:12<00:02, 6.04MB/s]
66%|######6 | 27.5M/41.5M [00:12<00:02, 7.27MB/s]
68%|######7 | 28.2M/41.5M [00:12<00:01, 7.13MB/s]
70%|######9 | 28.9M/41.5M [00:12<00:01, 6.61MB/s]
72%|#######2 | 30.1M/41.5M [00:12<00:01, 8.01MB/s]
74%|#######4 | 30.9M/41.5M [00:12<00:01, 7.87MB/s]
76%|#######6 | 31.6M/41.5M [00:12<00:01, 7.29MB/s]
79%|#######9 | 32.9M/41.5M [00:13<00:01, 8.81MB/s]
81%|########1 | 33.8M/41.5M [00:13<00:00, 8.65MB/s]
83%|########3 | 34.6M/41.5M [00:13<00:00, 8.01MB/s]
86%|########6 | 35.9M/41.5M [00:13<00:00, 8.85MB/s]
90%|#####
###9 | 37.3M/41.5M [00:13<00:00, 10.3MB/s]
92%|#########2| 38.3M/41.5M [00:13<00:00, 8.94MB/s]
94%|#########4| 39.2M/41.5M [00:13<00:00, 8.37MB/s]
97%|#########7| 40.3M/41.5M [00:13<00:00, 8.72MB/s]
100%|##########| 41.5M/41.5M [00:14<00:00, 3.11MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
0%| | 16.0k/41.5M [00:00<07:40, 94.3kB/s]
0%| | 48.0k/41.5M [00:00<04:51, 149kB/s]
0%| | 104k/41.5M [00:00<03:07, 231kB/s]
0%| | 208k/41.5M [00:00<01:53, 381kB/s]
1%| | 424k/41.5M [00:00<01:01, 700kB/s]
2%|2 | 864k/41.5M [00:01<00:31, 1.34MB/s]
4%|4 | 1.69M/41.5M [00:01<00:16, 2.56MB/s]
8%|7 | 3.16M/41.5M [00:01<00:08, 4.55MB/s]
11%|#1 | 4.62M/41.5M [00:01<00:06, 5.88MB/s]
15%|#4 | 6.09M/41.5M [00:01<00:05, 6.79MB/s]
18%|#8 | 7.56M/41.5M [00:01<00:04, 7.41MB/s]
22%|##1 | 9.04M/41.5M [00:02<00:04, 7.85MB/s]
25%|##5 | 10.5M/41.5M [00:02<00:03, 8.13MB/s]
29%|##8 | 12.0M/41.5M [00:02<00:03, 8.34MB/s]
32%|###2 | 13.4M/41.5M [00:02<00:03, 8.48MB/s]
36%|###5 | 14.9M/41.5M [00:02<00:03, 8.58MB/s]
39%|###9 | 16.4M/41.5M [00:02<00
:03, 8.64MB/s]
43%|####3 | 17.8M/41.5M [00:03<00:02, 8.69MB/s]
47%|####6 | 19.3M/41.5M [00:03<00:02, 8.73MB/s]
50%|##### | 20.8M/41.5M [00:03<00:02, 8.74MB/s]
54%|#####3 | 22.2M/41.5M [00:03<00:02, 8.76MB/s]
57%|#####7 | 23.7M/41.5M [00:03<00:02, 8.78MB/s]
61%|###### | 25.2M/41.5M [00:04<00:01, 8.79MB/s]
64%|######4 | 26.6M/41.5M [00:04<00:01, 8.79MB/s]
68%|######7 | 28.1M/41.5M [00:04<00:01, 8.79MB/s]
71%|#######1 | 29.6M/41.5M [00:04<00:01, 8.79MB/s]
75%|#######4 | 31.0M/41.5M [00:04<00:01, 8.80MB/s]
78%|#######8 | 32.5M/41.5M [00:04<00:01, 8.80MB/s]
82%|########1 | 34.0M/41.5M [00:05<00:00, 8.81MB/s]
85%|########5 | 35.5M/41.5M [00:05<00:00, 8.81MB/s]
89%|########8 | 36.9M/41.5M [00:05<00:00, 8.80MB/s]
93%|#########2| 38.4M/41.5M [00:05<00:00, 9.84MB/s]
96%|#########6| 39.8M/41.5M [00:05<00:00, 10.3MB/s]
98%|#########8| 40.9M/41.5M [00:05<00:00, 10.1MB/s]
100%|####
######| 41.5M/41.5M [00:05<00:00, 7.32MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_paddle.rst.txt b/docs/_sources/how_to/compile_models/from_paddle.rst.txt
index dde033c4d..6e6af7e6b 100644
--- a/docs/_sources/how_to/compile_models/from_paddle.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_paddle.rst.txt
@@ -210,7 +210,7 @@ Look up prediction top 1 index in 1000 class synset.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 10.497 seconds)
+ **Total running time of the script:** ( 1 minutes 6.877 seconds)
.. _sphx_glr_download_how_to_compile_models_from_paddle.py:
diff --git a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
index 0daa78d9f..6744d2794 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -79,7 +79,7 @@ Load a pretrained PyTorch model
.. code-block:: none
Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
-
0%| | 0.00/44.7M [00:00<?, ?B/s]
43%|####3 | 19.4M/44.7M [00:00<00:00, 203MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 238MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
44%|####3 | 19.6M/44.7M [00:00<00:00, 206MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 239MB/s]
diff --git a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
index a40dfa66c..abb555955 100644
--- a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
@@ -5,15 +5,15 @@
Computation times
=================
-**05:24.411** total execution time for **how_to_compile_models** files:
+**05:31.303** total execution time for **how_to_compile_models** files:
-- **01:10.497**: :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)
-- **00:59.180**: :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``)
-- **00:56.550**: :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)
-- **00:37.552**: :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)
-- **00:24.066**: :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)
-- **00:21.036**: :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)
-- **00:20.886**: :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)
-- **00:18.870**: :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)
-- **00:13.327**: :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)
-- **00:02.447**: :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)
+- **01:06.877**: :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)
+- **00:59.147**: :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``)
+- **00:57.219**: :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)
+- **00:39.405**: :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)
+- **00:30.243**: :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)
+- **00:21.431**: :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)
+- **00:20.708**: :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)
+- **00:19.861**: :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)
+- **00:13.896**: :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)
+- **00:02.516**: :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)
diff --git a/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt b/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
index 4ac8137f1..7b262f34f 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
@@ -402,7 +402,7 @@ Execute on TVM
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.4666 15.4579 15.6077 15.4158 0.0530
+ 16.0870 16.0837 16.1489 16.0308 0.0370
diff --git a/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt b/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
index fd6cb693f..3f6818da2 100644
--- a/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
@@ -108,7 +108,7 @@ Load pre-trained maskrcnn from torchvision and do tracing
.. code-block:: none
Downloading: "https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth" to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
-
0%| | 0.00/170M [00:00<?, ?B/s]
13%|#3 | 22.2M/170M [00:00<00:00, 233MB/s]
30%|##9 | 50.4M/170M [00:00<00:00, 270MB/s]
45%|####4 | 76.1M/170M [00:00<00:00, 255MB/s]
60%|#####9 | 101M/170M [00:00<00:00, 259MB/s]
76%|#######6 | 130M/170M [00:00<00:00, 272MB/s]
93%|#########2| 158M/170M [00:00<00:00, 280MB/s]
100%|##########| 170M/170M [00:00<00:00, 272MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
3%|2 | 4.44M/170M [00:00<00:03, 46.4MB/s]
5%|5 | 8.87M/170M [00:00<00:03, 45.5MB/s]
22%|##1 | 36.5M/170M [00:00<00:00, 156MB/s]
38%|###8 | 65.0M/170M [00:00<00:00, 212MB/s]
55%|#####5 | 93.4M/170M [00:00<00:00, 243MB/s]
72%|#######1 | 122M/170M [00:00<00:00, 262MB/s]
89%|########8 | 150M/170M [00:00<00:00, 273MB/s]
100%|##########| 170M/170M [00:00<00:00, 231MB/s]
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:3878: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
for i in range(dim)
/usr/local/lib/python3.7/dist-packages/torchvision/models/detection/anchor_utils.py:127: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
@@ -262,7 +262,7 @@ Get boxes with score larger than 0.9
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 56.734 seconds)
+ **Total running time of the script:** ( 3 minutes 1.543 seconds)
.. _sphx_glr_download_how_to_deploy_models_deploy_object_detection_pytorch.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
index 6522b397a..a82697969 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
@@ -187,7 +187,7 @@ training. Other models require a full post training calibration.
.. code-block:: none
Downloading: "https://download.pytorch.org/models/mobilenet_v2-b0353104.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
-
0%| | 0.00/13.6M [00:00<?, ?B/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 182MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 185MB/s]
@@ -353,7 +353,7 @@ Here we give an example of how to measure performance of TVM compiled models.
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 90.5653 90.6645 91.4601 90.0798 0.2444
+ 90.3016 90.2044 90.8585 89.8834 0.2526
@@ -393,7 +393,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 3.979 seconds)
+ **Total running time of the script:** ( 1 minutes 4.650 seconds)
.. _sphx_glr_download_how_to_deploy_models_deploy_prequantized.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt b/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
index cde7b0bf7..67fda532a 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
@@ -360,7 +360,7 @@ Here we give an example of how to measure performance of TVM compiled models.
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 116.9583 116.7426 119.2221 115.7829 0.8334
+ 120.5670 120.5832 121.4224 119.8656 0.2797
@@ -394,7 +394,7 @@ Here we give an example of how to measure performance of TVM compiled models.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 56.558 seconds)
+ **Total running time of the script:** ( 1 minutes 56.617 seconds)
.. _sphx_glr_download_how_to_deploy_models_deploy_prequantized_tflite.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
index 2bc61aa8c..db4147810 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -223,7 +223,7 @@ We create a Relay VM to build and execute the model.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 11.625 seconds)
+ **Total running time of the script:** ( 2 minutes 14.102 seconds)
.. _sphx_glr_download_how_to_deploy_models_deploy_quantized.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt b/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
index 72ec5932c..4c8e6a2b6 100644
--- a/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
@@ -137,7 +137,7 @@ Convert and compile model for CPU.
data: None
input_sym_arg_type = in_param.infer_type()[0]
Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
-
0%| | 0/132723 [00:00<?, ?KB/s]
5%|5 | 7046/132723 [00:00<00:01, 70452.51KB/s]
12%|#1 | 15578/132723 [00:00<00:01, 79189.87KB/s]
18%|#8 | 24179/132723 [00:00<00:01, 82297.79KB/s]
25%|##4 | 32822/132723 [00:00<00:01, 83926.10KB/s]
31%|###1 | 41433/132723 [00:00<00:01, 84712.06KB/s]
38%|###7 | 50032/132723 [00:00<00:00, 85139.87KB/s]
44%|####4 | 58637/132723 [00:00<00:00, 85433.22KB/s]
51%|##### | 67181/132723 [00:00<00:00, 85428.20KB/s]
57%|#####7 | 75791/132723 [00:00<00:00, 85635.41KB/s]
64%|######3 | 84355/132723 [00:01<00:00, 85524.31KB/s]
70%|####### | 92908/132723 [00:01<00:00, 85273.26KB/s]
76%|#######6 | 101436/132723 [00:01<00:00, 85171.71KB/s]
83%|########2 | 109954/132723 [00:01<00:00, 85144.31KB/s]
89%|########9 | 118469/132723 [00:01<00:00, 84959.97KB/s]
96%|#########5| 126975/132723 [00:01<00:00, 84989.14KB/s]
100%|#######
###| 132723/132723 [00:01<00:00, 84426.01KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
5%|4 | 6422/132723 [00:00<00:01, 64211.95KB/s]
11%|#1 | 15033/132723 [00:00<00:01, 77090.48KB/s]
18%|#7 | 23723/132723 [00:00<00:01, 81565.56KB/s]
24%|##4 | 32392/132723 [00:00<00:01, 83583.28KB/s]
31%|### | 41113/132723 [00:00<00:01, 84880.23KB/s]
38%|###7 | 49818/132723 [00:00<00:00, 85616.43KB/s]
44%|####4 | 58477/132723 [00:00<00:00, 85932.03KB/s]
51%|##### | 67168/132723 [00:00<00:00, 86238.65KB/s]
57%|#####7 | 75937/132723 [00:00<00:00, 86688.92KB/s]
64%|######3 | 84687/132723 [00:01<00:00, 86937.59KB/s]
70%|####### | 93422/132723 [00:01<00:00, 87061.35KB/s]
77%|#######6 | 102196/132723 [00:01<00:00, 87265.66KB/s]
84%|########3 | 110927/132723 [00:01<00:00, 87276.34KB/s]
90%|######### | 119696/132723 [00:01<00:00, 87399.54KB/s]
97%|#########6| 128436/132723 [00:01<00:00, 87391.35KB/s]
100%|#######
###| 132723/132723 [00:01<00:00, 85534.95KB/s]
@@ -211,7 +211,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 35.845 seconds)
+ **Total running time of the script:** ( 2 minutes 37.568 seconds)
.. _sphx_glr_download_how_to_deploy_models_deploy_ssd_gluoncv.py:
diff --git a/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt b/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
index 66a01e560..40ea0e97d 100644
--- a/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
@@ -5,13 +5,13 @@
Computation times
=================
-**10:34.413** total execution time for **how_to_deploy_models** files:
+**11:44.196** total execution time for **how_to_deploy_models** files:
-- **02:56.734**: :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``)
-- **02:35.845**: :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)
-- **01:56.558**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)
-- **01:11.625**: :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)
-- **01:03.979**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)
-- **00:27.872**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)
-- **00:21.621**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)
-- **00:00.177**: :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``)
+- **03:01.543**: :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``)
+- **02:37.568**: :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)
+- **02:14.102**: :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)
+- **01:56.617**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)
+- **01:04.650**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)
+- **00:27.652**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)
+- **00:21.858**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)
+- **00:00.206**: :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``)
diff --git a/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt b/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
index d006b5e3b..168debf8b 100644
--- a/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
@@ -425,7 +425,7 @@ First let us define two helper functions to get the mobilenet model and a cat im
.. code-block:: none
- Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip448c3c6b-e2a1-468b-abfd-6265e0f34ccf from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip7a20ba62-ec3d-4d09-90ea-6fe5c24e8cab from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
diff --git a/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt b/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
index 88fdb021f..31aa377c0 100644
--- a/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
@@ -5,9 +5,9 @@
Computation times
=================
-**00:37.164** total execution time for **how_to_extend_tvm** files:
+**00:37.575** total execution time for **how_to_extend_tvm** files:
-- **00:33.792**: :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``)
-- **00:02.180**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)
-- **00:01.009**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)
-- **00:00.183**: :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``)
+- **00:34.140**: :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``)
+- **00:02.204**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)
+- **00:01.023**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)
+- **00:00.207**: :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``)
diff --git a/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt b/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
index fc8431e22..4f78aad77 100644
--- a/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
@@ -199,10 +199,10 @@ profile the execution time of each passes.
.. code-block:: none
Printing results of timing profile...
- InferType: 6103us [6103us] (45.64%; 45.64%)
- FoldScaleAxis: 7269us [6us] (54.36%; 54.36%)
- FoldConstant: 7263us [1477us] (54.32%; 99.92%)
- InferType: 5786us [5786us] (43.27%; 79.66%)
+ InferType: 6028us [6028us] (45.49%; 45.49%)
+ FoldScaleAxis: 7222us [6us] (54.51%; 54.51%)
+ FoldConstant: 7216us [1473us] (54.46%; 99.92%)
+ InferType: 5744us [5744us] (43.35%; 79.59%)
@@ -239,10 +239,10 @@ Refer to following sections and :py:func:`tvm.instrument.pass_instrument` for th
.. code-block:: none
Printing results of timing profile...
- InferType: 5825us [5825us] (44.52%; 44.52%)
- FoldScaleAxis: 7257us [5us] (55.48%; 55.48%)
- FoldConstant: 7253us [1517us] (55.44%; 99.94%)
- InferType: 5736us [5736us] (43.85%; 79.09%)
+ InferType: 5810us [5810us] (44.63%; 44.63%)
+ FoldScaleAxis: 7208us [5us] (55.37%; 55.37%)
+ FoldConstant: 7203us [1510us] (55.33%; 99.94%)
+ InferType: 5694us [5694us] (43.74%; 79.04%)
diff --git a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
index 60e5d0540..9814b1469 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
@@ -295,7 +295,7 @@ latency of convolution.
.. code-block:: none
- Convolution: 54.171636 ms
+ Convolution: 37.823921 ms
diff --git a/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt b/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
index 57061b4e2..8472f80fb 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
@@ -628,7 +628,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 6.863368 ms
+ conv2d with tensor core: 9.411514 ms
diff --git a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
index a8d340cf0..54f93d0ae 100644
--- a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
@@ -118,8 +118,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
.. code-block:: none
- Numpy running time: 0.018139
- Baseline: 3.389786
+ Numpy running time: 0.018413
+ Baseline: 3.375813
@@ -210,7 +210,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.291762
+ Opt1: 0.293211
@@ -309,7 +309,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.331351
+ Opt2: 0.330585
@@ -401,7 +401,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.113804
+ Opt3: 0.113394
@@ -520,7 +520,7 @@ flattening.
.. code-block:: none
- Opt4: 0.109979
+ Opt4: 0.109488
@@ -638,7 +638,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.111204
+ Opt5: 0.110785
@@ -759,7 +759,7 @@ Futhermore, we can also utilize multi-core processors to do the thread-level par
.. code-block:: none
- Opt6: 0.144465
+ Opt6: 0.143222
diff --git a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
index c371be56f..308ff14d7 100644
--- a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
@@ -5,8 +5,8 @@
Computation times
=================
-**00:34.368** total execution time for **how_to_optimize_operators** files:
+**00:34.320** total execution time for **how_to_optimize_operators** files:
-- **00:32.023**: :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)
-- **00:01.266**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``)
-- **00:01.079**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)
+- **00:31.967**: :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)
+- **00:01.303**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``)
+- **00:01.051**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
index d4bc7cfa4..1fcfa6cc7 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
@@ -5,11 +5,11 @@
Computation times
=================
-**03:28.794** total execution time for **how_to_tune_with_autoscheduler** files:
-
-- **01:17.695**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)
-- **01:05.687**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``)
-- **00:39.971**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)
-- **00:08.592**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)
-- **00:08.541**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)
-- **00:08.308**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)
+**03:29.523** total execution time for **how_to_tune_with_autoscheduler** files:
+
+- **01:19.638**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)
+- **01:02.355**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``)
+- **00:40.302**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)
+- **00:10.386**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)
+- **00:08.549**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)
+- **00:08.292**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
index 85147c598..8e89909da 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
@@ -221,12 +221,12 @@ cooperative fetching, unrolling and operator fusion.
compute: Buffer(compute_2: Pointer(float32), float32, [25088], [])}
buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute}
preflattened_buffer_map = {data_1: data_3: Buffer(data_2, float32, [1, 512, 7, 7], []), kernel_1: kernel_3: Buffer(kernel_2, float32, [512, 512, 3, 3], []), bias_1: bias_3: Buffer(bias_2, float32, [1, 512, 1, 1], []), compute_1: compute_3: Buffer(compute_2, float32, [1, 512, 7, 7], [])} {
- attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 64;
- allocate(conv2d_nchw: Pointer(local float32), float32, [8]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [324]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [288]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [8], [], scope="local", align=32)[0] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 28;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [14], [], scope="local", align=32)[0] = 0f32
conv2d_nchw_1[1] = 0f32
conv2d_nchw_1[2] = 0f32
conv2d_nchw_1[3] = 0f32
@@ -234,88 +234,470 @@ cooperative fetching, unrolling and operator fusion.
conv2d_nchw_1[5] = 0f32
conv2d_nchw_1[6] = 0f32
conv2d_nchw_1[7] = 0f32
- for (rc.outer.outer: int32, 0, 128) {
- let cse_var_1: int32 = (rc.outer.outer*196)
- {
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [324], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((9 <= threadIdx.x_1) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[(((cse_var_1 + (floordiv(threadIdx.x_1, 9)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 49), 81)) && (floormod((threadIdx.x_1 + 49), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 49), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 49), 81), 9)*7)) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 <= floormod((threadIdx.x_1 + 8), 9)) && (floormod((threadIdx.x_1 + 8), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 98), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 98), 81), 9)*7)) + floormod((threadIdx.x_1 + 8), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 147), 81)) && (floormod((threadIdx.x_1 + 66), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 147), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 147), 81), 9)*7)) + floormod((threadIdx.x_1 + 3), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 196), 81)) && (floormod((threadIdx.x_1 + 34), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 7), 9))) && (floormod((threadIdx.x_1 + 7), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 196), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 196), 81), 9)*7)) + floormod((threadIdx.x_1 + 7), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((9 <= floormod((threadIdx.x_1 + 245), 81)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 245), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 245), 81), 9)*7)) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- if @tir.likely((threadIdx.x_1 < 30), dtype=bool) {
- pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((floormod((threadIdx.x_1 + 51), 81) < 72) && (1 <= floormod((threadIdx.x_1 + 6), 9))) && (floormod((threadIdx.x_1 + 6), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 294), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 294), 81), 9)*7)) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
- }
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1: Buffer(kernel.shared, float32, [288], [], scope="shared")[(threadIdx.x_2*8)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod((threadIdx.x_2*8), 36), 3)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 1)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 1), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 2)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 2), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 3)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floormod((floordiv((threadIdx.x_2*8), 3) + 1), 12)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 4)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 4), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 5)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 5), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 6)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floormod((floordiv((threadIdx.x_2*8), 3) + 2), 12)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 7)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 7), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- }
- for (rx.outer.inner: int32, 0, 3) {
- for (ff.outer.inner: int32, 0, 4) {
- let cse_var_4: int32 = (ff.outer.inner*2)
- let cse_var_3: int32 = ((ff.outer.inner*72) + rx.outer.inner)
- let cse_var_2: int32 = (cse_var_4 + 1)
- {
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[cse_var_3]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_3 + 36)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[(cse_var_3 + 3)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[(cse_var_3 + 39)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[(cse_var_3 + 6)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[(cse_var_3 + 42)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[(cse_var_3 + 9)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[(cse_var_3 + 45)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[(cse_var_3 + 12)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[(cse_var_3 + 48)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[(cse_var_3 + 15)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[(cse_var_3 + 51)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[(cse_var_3 + 18)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[(cse_var_3 + 54)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[(cse_var_3 + 21)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[(cse_var_3 + 57)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[(cse_var_3 + 24)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[(cse_var_3 + 60)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[(cse_var_3 + 27)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[(cse_var_3 + 63)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[(cse_var_3 + 30)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[(cse_var_3 + 66)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[(cse_var_3 + 33)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[(cse_var_3 + 69)]))
+ conv2d_nchw_1[8] = 0f32
+ conv2d_nchw_1[9] = 0f32
+ conv2d_nchw_1[10] = 0f32
+ conv2d_nchw_1[11] = 0f32
+ conv2d_nchw_1[12] = 0f32
+ conv2d_nchw_1[13] = 0f32
+ for (rc.outer.outer: int32, 0, 64) {
+ for (ry.outer.outer: int32, 0, 3) {
+ let cse_var_2: int32 = (rc.outer.outer*72)
+ let cse_var_1: int32 = (ry.outer.outer*3)
+ {
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope="shared")[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1*4), 9))) && (floormod((threadIdx.x_1*4), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv((threadIdx.x_1*4), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod((threadIdx.x_1*4), 9)) - 8)], 0f3 [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 1), 9))) && (floormod(((threadIdx.x_1*4) + 1), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 1), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 1), 9)) - 8)], 0f32, dtype=float32)
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 2), 9))) && (floormod(((threadIdx.x_1*4) + 2), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 2), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 2), 9)) - 8)], 0f32, dtype=float32)
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 3), 9))) && (floormod(((threadIdx.x_1*4) + 3), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 3), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 3), 9)) - 8)], 0f32, dtype=float32)
}
}
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope="shared")[threadIdx.x_2] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 64)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 8), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 128)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 16), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 32), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 192)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 36864)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 256)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 32), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 64), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 320)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 40), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 80), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 384)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 73728)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 448)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 56), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 112), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 512)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 64), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 128), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 576)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 110592)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 640)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 80), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 160), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 704)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 88), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 176), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 768)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 147456)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 832)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 104), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 208), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 896)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 112), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 224), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 960)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 184320)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 128), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 256), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 136), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 272), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 221184)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 152), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 304), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 160), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 320), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 258048)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 176), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 352), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 184), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 368), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 294912)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 200), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 400), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 208), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 416), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 331776)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 224), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 448), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 232), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 464), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 368640)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 248), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 496), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 256), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 512), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 405504)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 272), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 544), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 280), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 560), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 442368)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 296), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 592), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 304), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 608), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 479232)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 320), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 640), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 328), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 656), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 516096)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 344), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 688), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 352), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 704), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 552960)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 368), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 736), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 376), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 752), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 47)]))
}
}
}
- for (i1.inner: int32, 0, 8) {
- compute[(((blockIdx.x*392) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias[((blockIdx.x*8) + i1.inner)]), 0f32)
+ for (i1.inner: int32, 0, 2) {
+ for (i3.inner: int32, 0, 7) {
+ compute[(((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*98)) + (i1.inner*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[((i1.inner*7) + i3.inner)] + bias[(((floordiv(blockIdx.x, 7)*128) + (threadIdx.x*2)) + i1.inner)]), 0f32)
+ }
}
}
}
@@ -368,7 +750,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.311 ms
+ Execution time of this operator: 0.365 ms
@@ -412,21 +794,21 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
- conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
- conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=4)
- conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=1)
+ conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
+ conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+ conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=64)
conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
- conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+ conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
- conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
- conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=7)
+ conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=7)
+ conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
- conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=4)
- conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
- conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
+ conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=2)
+ conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
+ conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
@@ -434,14 +816,14 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
- compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=8)
- compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=1)
+ compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
+ compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=64)
compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
- compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+ compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
- compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
- compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
+ compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
+ compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
@@ -459,16 +841,16 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused = s[compute].fuse(compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i)
s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread_axis("threadIdx.x"))
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
- kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=8)
+ kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
- kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
+ kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
- pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
- pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
+ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
- s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 64)
+ s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -486,10 +868,10 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
#define int64_t long long
#define uint64_t unsigned long long
#endif
- extern "C" __global__ void __launch_bounds__(49) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[8];
- __shared__ float pad_temp_shared[324];
- __shared__ float kernel_shared[288];
+ extern "C" __global__ void __launch_bounds__(64) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[14];
+ __shared__ float pad_temp_shared[72];
+ __shared__ float kernel_shared[3072];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
conv2d_nchw[2] = 0.000000e+00f;
@@ -498,73 +880,418 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[7] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
- __syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = ((((9 <= ((int)threadIdx.x)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((9 <= ((((int)threadIdx.x) + 49) % 81)) && (((((int)threadIdx.x) + 49) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 49) / 81) * 49)) + ((((((int)threadIdx.x) + 49) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 <= ((((int)threadIdx.x) + 8) % 9)) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 98) / 81) * 49)) + ((((((int)threadIdx.x) + 17) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((9 <= ((((int)threadIdx.x) + 66) % 81)) && (((((int)threadIdx.x) + 66) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 147) / 81) * 49)) + ((((((int)threadIdx.x) + 66) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((9 <= ((((int)threadIdx.x) + 34) % 81)) && (((((int)threadIdx.x) + 34) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 196) / 81) * 49)) + ((((((int)threadIdx.x) + 34) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((9 <= ((((int)threadIdx.x) + 2) % 81)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 245) / 81) * 49)) + ((((((int)threadIdx.x) + 2) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
- if (((int)threadIdx.x) < 30) {
- pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((((int)threadIdx.x) < 21) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 294) / 81) * 49)) + ((((((int)threadIdx.x) + 51) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[(((int)threadIdx.x) * 8)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) * 8) % 36) / 3) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 1)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 1) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 2)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 2) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 3)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) / 3) + 1) % 12) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 4)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 4) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 5)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 5) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 6)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) / 3) + 2) % 12) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 7)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 7) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- __syncthreads();
- for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
- for (int ff_outer_inner = 0; ff_outer_inner < 4; ++ff_outer_inner) {
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[((ff_outer_inner * 72) + rx_outer_inner)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 36)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 3)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 39)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 6)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 42)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 9)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 45)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 12)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 48)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 15)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 51)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 18)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 54)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 21)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 57)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 24)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 60)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 27)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 63)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 30)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 66)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 33)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 69)]));
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[13] = 0.000000e+00f;
+ for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
+ for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
+ __syncthreads();
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) * 4) % 9))) && (((((int)threadIdx.x) * 4) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 4) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 1) % 9))) && ((((((int)threadIdx.x) * 4) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.000000e+00f);
}
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 2) % 9))) && ((((((int)threadIdx.x) * 4) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 3) % 9))) && ((((((int)threadIdx.x) * 4) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.000000e+00f);
+ }
+ kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 36864)];
+ kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 73728)];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 110592)];
+ kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 147456)];
+ kernel_shared[(((int)threadIdx.x) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 184320)];
+ kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 221184)];
+ kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 258048)];
+ kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 294912)];
+ kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 331776)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 368640)];
+ kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 405504)];
+ kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 442368)];
+ kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 479232)];
+ kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 516096)];
+ kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 552960)];
+ kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ __syncthreads();
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
}
}
- for (int i1_inner = 0; i1_inner < 8; ++i1_inner) {
- compute[(((((int)blockIdx.x) * 392) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 8) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[((((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[((((((int)blockIdx.x) / 7) * 128) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+ }
}
}
@@ -623,7 +1350,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 5.687 seconds)
+ **Total running time of the script:** ( 1 minutes 2.355 seconds)
.. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py:
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
index 0c9fab1c5..4ef4603b9 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
@@ -616,7 +616,7 @@ so we can read the log file and load the best schedules.
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 9.9082 9.9383 9.9670 9.8194 0.0639
+ 9.9023 9.8786 9.9577 9.8707 0.0392
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
index bd105b521..60f7f7cb7 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
@@ -635,7 +635,7 @@ so we can read the log file and load the best schedules.
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 749.4805 749.3528 750.0807 749.0080 0.4471
+ 804.0846 803.9296 804.5431 803.7811 0.3298
@@ -660,7 +660,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 17.695 seconds)
+ **Total running time of the script:** ( 1 minutes 19.638 seconds)
.. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_network_x86.py:
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
index cd040d3c5..ab1fffc6c 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
@@ -361,71 +361,69 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [65536], []),
compute: Buffer(compute_2: Pointer(float32), float32, [65536], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute}
- preflattened_buffer_map = {placeholder_5: placeholder_15: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], []), placeholder_6: placeholder_16: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_9: placeholder_17: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_18: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_19: Buffer(placeholder_12, int32, [4916], [])} {
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_9: placeholder_16: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_5: placeholder_19: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
for (i0.outer.i1.outer.fused: int32, 0, 32) "parallel" {
allocate(compute_4: Pointer(global float32), float32, [2048]), storage_scope = global {
- for (i.outer.inner: int32, 0, 2) {
- for (nb_j.inner: int32, 0, 2) {
- for (i.inner.init: int32, 0, 32) {
- let cse_var_1: int32 = (((i.outer.inner*1024) + (i.inner.init*32)) + (nb_j.inner*16))
- {
- compute_5: Buffer(compute_4, float32, [2048], [])[cse_var_1] = 0f32
- compute_5[(cse_var_1 + 1)] = 0f32
- compute_5[(cse_var_1 + 2)] = 0f32
- compute_5[(cse_var_1 + 3)] = 0f32
- compute_5[(cse_var_1 + 4)] = 0f32
- compute_5[(cse_var_1 + 5)] = 0f32
- compute_5[(cse_var_1 + 6)] = 0f32
- compute_5[(cse_var_1 + 7)] = 0f32
- compute_5[(cse_var_1 + 8)] = 0f32
- compute_5[(cse_var_1 + 9)] = 0f32
- compute_5[(cse_var_1 + 10)] = 0f32
- compute_5[(cse_var_1 + 11)] = 0f32
- compute_5[(cse_var_1 + 12)] = 0f32
- compute_5[(cse_var_1 + 13)] = 0f32
- compute_5[(cse_var_1 + 14)] = 0f32
- compute_5[(cse_var_1 + 15)] = 0f32
- }
+ for (nb_j.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 64) {
+ let cse_var_1: int32 = ((i.inner.init*32) + (nb_j.inner*16))
+ {
+ compute_5: Buffer(compute_4, float32, [2048], [])[cse_var_1] = 0f32
+ compute_5[(cse_var_1 + 1)] = 0f32
+ compute_5[(cse_var_1 + 2)] = 0f32
+ compute_5[(cse_var_1 + 3)] = 0f32
+ compute_5[(cse_var_1 + 4)] = 0f32
+ compute_5[(cse_var_1 + 5)] = 0f32
+ compute_5[(cse_var_1 + 6)] = 0f32
+ compute_5[(cse_var_1 + 7)] = 0f32
+ compute_5[(cse_var_1 + 8)] = 0f32
+ compute_5[(cse_var_1 + 9)] = 0f32
+ compute_5[(cse_var_1 + 10)] = 0f32
+ compute_5[(cse_var_1 + 11)] = 0f32
+ compute_5[(cse_var_1 + 12)] = 0f32
+ compute_5[(cse_var_1 + 13)] = 0f32
+ compute_5[(cse_var_1 + 14)] = 0f32
+ compute_5[(cse_var_1 + 15)] = 0f32
}
- for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
- for (i.inner: int32, 0, 32) {
- let cse_var_21: int32 = (elem_idx*16)
- let cse_var_20: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
- let cse_var_19: int32 = (((i.outer.inner*1024) + (i.inner*32)) + (nb_j.inner*16))
- let cse_var_18: int32 = (cse_var_19 + 1)
- let cse_var_17: int32 = (cse_var_19 + 11)
- let cse_var_16: int32 = (cse_var_19 + 12)
- let cse_var_15: int32 = (cse_var_19 + 13)
- let cse_var_14: int32 = (cse_var_19 + 14)
- let cse_var_13: int32 = (cse_var_19 + 15)
- let cse_var_12: int32 = (cse_var_19 + 2)
- let cse_var_11: int32 = (cse_var_19 + 3)
- let cse_var_10: int32 = (cse_var_19 + 4)
- let cse_var_9: int32 = (cse_var_19 + 5)
- let cse_var_8: int32 = (cse_var_19 + 6)
- let cse_var_7: int32 = (cse_var_19 + 7)
- let cse_var_6: int32 = (cse_var_19 + 8)
- let cse_var_5: int32 = (cse_var_19 + 9)
- let cse_var_4: int32 = (((floordiv(i0.outer.i1.outer.fused, 16)*16384) + (i.outer.inner*8192)) + (i.inner*256))
- let cse_var_3: int32 = (cse_var_19 + 10)
- {
- compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_20]*16) + cse_var_21)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 1)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 2)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 3)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 4)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 5)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 6)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 7)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 8)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 9)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 10)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 11)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 12)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 13)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 14)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 15)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- }
+ }
+ for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
+ for (i.inner: int32, 0, 64) {
+ let cse_var_21: int32 = (elem_idx*16)
+ let cse_var_20: int32 = ((i.inner*32) + (nb_j.inner*16))
+ let cse_var_19: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
+ let cse_var_18: int32 = (cse_var_20 + 1)
+ let cse_var_17: int32 = (cse_var_20 + 11)
+ let cse_var_16: int32 = (cse_var_20 + 12)
+ let cse_var_15: int32 = (cse_var_20 + 13)
+ let cse_var_14: int32 = (cse_var_20 + 14)
+ let cse_var_13: int32 = (cse_var_20 + 15)
+ let cse_var_12: int32 = (cse_var_20 + 2)
+ let cse_var_11: int32 = (cse_var_20 + 3)
+ let cse_var_10: int32 = (cse_var_20 + 4)
+ let cse_var_9: int32 = (cse_var_20 + 5)
+ let cse_var_8: int32 = (cse_var_20 + 6)
+ let cse_var_7: int32 = (cse_var_20 + 7)
+ let cse_var_6: int32 = (cse_var_20 + 8)
+ let cse_var_5: int32 = (cse_var_20 + 9)
+ let cse_var_4: int32 = ((floordiv(i0.outer.i1.outer.fused, 16)*16384) + (i.inner*256))
+ let cse_var_3: int32 = (cse_var_20 + 10)
+ {
+ compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[((placeholder_3[cse_var_19]*16) + cse_var_21)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 1)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 2)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 3)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 4)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 5)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 6)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 7)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 8)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 9)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 10)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 11)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 12)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 13)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 14)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 15)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
}
}
}
@@ -486,7 +484,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 1.724 ms
+ Execution time of this operator: 1.837 ms
diff --git a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
index 20874d9fc..f00df1b4a 100644
--- a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
Computation times
=================
-**00:45.186** total execution time for **how_to_tune_with_autotvm** files:
+**00:45.261** total execution time for **how_to_tune_with_autotvm** files:
-- **00:44.398**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)
-- **00:00.206**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)
-- **00:00.196**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)
-- **00:00.194**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``)
-- **00:00.191**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.py``)
+- **00:44.454**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)
+- **00:00.219**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.py``)
+- **00:00.204**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)
+- **00:00.192**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``)
+- **00:00.191**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)
diff --git a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
index 827f86c8c..871d7d87e 100644
--- a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
@@ -859,8 +859,8 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2885496
- No: 6 GFLOPS: 111.00/111.00 result: MeasureResult(costs=(0.002085639479166667,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.2219181060791016, timestamp=1654041265.559902) [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3754080
- No: 7 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 6 GFLOPS: 109.94/109.94 result: MeasureResult(costs=(0.0021056749375,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.230674982070923, timestamp=1654091865.1226287) [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3754080
+ No: 7 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -983,7 +983,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 16, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6225319
- No: 8 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 8 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1106,7 +1106,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,943546
- No: 9 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 9 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1229,7 +1229,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2868708
- No: 10 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 10 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
res = future.result()
File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
@@ -1247,7 +1247,7 @@ for this template
TimeoutError
[('tile_f', [-1, 32, 2, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4691833
- No: 11 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 11 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1370,7 +1370,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 2, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1042124
- No: 12 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 12 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1493,7 +1493,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10013405
- No: 13 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 13 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1616,7 +1616,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 8, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6732082
- No: 14 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 14 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1739,7 +1739,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7536735
- No: 15 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 15 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1862,7 +1862,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,482121
- No: 16 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 16 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1985,7 +1985,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2824525
- No: 17 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 17 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2108,7 +2108,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4559286
- No: 18 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 18 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2231,7 +2231,7 @@ for this template
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9677544
- No: 19 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+ No: 19 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 721, in __call__
yield remote, remote.load_module(os.path.split(build_result.filename)[1])
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 685, in run_through_rpc
@@ -2319,7 +2319,7 @@ for this template
15: _PyEval_EvalFrameDefault
14: 0x0000000000537c30
13: _PyObject_FastCallKeywords
- 12: 0x00007f341f4e3fa2
+ 12: 0x00007f1ae6d0bfa2
11: _ctypes_callproc
10: ffi_call
9: ffi_call_unix64
@@ -2384,7 +2384,7 @@ for this template
21: _PyFunction_FastCallKeywords
20: _PyEval_EvalFrameDefault
19: _PyFunction_FastCall [('tile_f', [-1, 8, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6390073
- No: 20 GFLOPS: 143.96/143.96 result: MeasureResult(costs=(0.00160804004,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.390920877456665, timestamp=1654041283.8257668) [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
+ No: 20 GFLOPS: 143.88/143.88 result: MeasureResult(costs=(0.0016089895300000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3828744888305664, timestamp=1654091884.3543549) [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
@@ -2437,7 +2437,7 @@ and measure running time.
Best config:
[('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
- Time cost of this operator: 0.001975
+ Time cost of this operator: 0.001973
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
index 52b1d8bf4..b02c498e8 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
@@ -294,10 +294,10 @@ Timing the untuned program
########## Build without Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs
--------- --- -------- ------- ----- ------ -------
- tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 311.5 98.763 (1, 2, 10, 10, 3) 2 1
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.951 (1, 6, 10, 10) 1 1
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 0.286 (1, 1, 10, 10, 3) 1 1
- Total_time - 315.401 - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.1 98.758 (1, 2, 10, 10, 3) 2 1
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.955 (1, 6, 10, 10) 1 1
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 0.287 (1, 1, 10, 10, 3) 1 1
+ Total_time - 314.001 - - - -
@@ -359,10 +359,10 @@ Timing the tuned program
########## Build with Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs
--------- --- -------- ------- ----- ------ -------
- tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 198.7 98.744 (1, 6, 10, 10, 1) 2 1
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.728 0.859 (1, 6, 10, 10) 1 1
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.8 0.398 (1, 3, 10, 10, 1) 1 1
- Total_time - 201.228 - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 78.45 96.779 (1, 6, 10, 10, 1) 2 1
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.71 2.11 (1, 6, 10, 10) 1 1
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 1.111 (1, 1, 10, 10, 3) 1 1
+ Total_time - 81.061 - - - -
diff --git a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
index 0a0d0032f..db2281f02 100644
--- a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
Computation times
=================
-**00:45.415** total execution time for **how_to_work_with_microtvm** files:
+**00:46.313** total execution time for **how_to_work_with_microtvm** files:
-- **00:41.463**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)
-- **00:03.403**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)
-- **00:00.189**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.py``)
-- **00:00.183**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.py``)
+- **00:42.355**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)
+- **00:03.406**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)
+- **00:00.190**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.py``)
+- **00:00.185**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.py``)
- **00:00.177**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_reference_vm.py` (``micro_reference_vm.py``)
diff --git a/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
index 42761cf89..600f50c06 100644
--- a/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
@@ -5,8 +5,8 @@
Computation times
=================
-**00:05.015** total execution time for **how_to_work_with_relay** files:
+**00:05.058** total execution time for **how_to_work_with_relay** files:
-- **00:03.423**: :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)
-- **00:01.381**: :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)
-- **00:00.210**: :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``)
+- **00:03.438**: :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)
+- **00:01.408**: :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)
+- **00:00.212**: :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``)
diff --git a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
index ddb7acf71..a84739cfa 100644
--- a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
@@ -5,13 +5,13 @@
Computation times
=================
-**00:04.940** total execution time for **how_to_work_with_schedules** files:
+**00:04.922** total execution time for **how_to_work_with_schedules** files:
-- **00:01.864**: :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)
-- **00:00.773**: :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)
-- **00:00.669**: :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)
-- **00:00.664**: :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)
-- **00:00.302**: :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)
-- **00:00.233**: :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``)
-- **00:00.224**: :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)
-- **00:00.210**: :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)
+- **00:01.844**: :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)
+- **00:00.764**: :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)
+- **00:00.666**: :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)
+- **00:00.657**: :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)
+- **00:00.301**: :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)
+- **00:00.239**: :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``)
+- **00:00.232**: :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)
+- **00:00.219**: :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)
diff --git a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
index c351debf6..e38561d2c 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -318,7 +318,7 @@ The importing needs to happen before the tensorized GEMV being executed.
C: Buffer(C_2: Pointer(float32), float32, [524288], [])}
buffer_map = {A_1: A, B_1: B, C_1: C}
preflattened_buffer_map = {A_1: A_3: Buffer(A_2, float32, [1024, 64], []), B_1: B_3: Buffer(B_2, float32, [512, 64], []), C_1: C_3: Buffer(C_2, float32, [1024, 512], [])} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpoxih48lj/input0.cc'\nsource_filename = \"/tmp/tmpoxih48lj/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca float*, align 8\n %8 = alloca float*, align 8\n %9 = alloca floa [...]
+ attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmp3nlbt5gp/input0.cc'\nsource_filename = \"/tmp/tmp3nlbt5gp/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = alloca float*, align 8\n %8 = alloca float*, align 8\n %9 = alloca floa [...]
for (i, 0, 1024) {
for (j.outer: int32, 0, 32) {
@tir.call_extern("gemv_update", @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
index cc0b842a2..b752a361f 100644
--- a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
@@ -5,7 +5,7 @@
Computation times
=================
-**00:20.285** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:21.533** total execution time for **topic_vta_tutorials_autotvm** files:
-- **00:20.077**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``)
-- **00:00.208**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)
+- **00:21.316**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``)
+- **00:00.217**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
index b6c782588..957f5be28 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -267,7 +267,7 @@ The compilation steps are:
DeprecationWarning,
/workspace/vta/tutorials/frontend/deploy_classification.py:213: DeprecationWarning: legacy graph executor behavior of producing json / lib / params will be removed in the next release. Please see documents of tvm.contrib.graph_executor.GraphModule for the new recommended usage.
relay_prog, target=tvm.target.Target(target, host=env.target_host), params=params
- resnet18_v1 inference graph built in 20.97s!
+ resnet18_v1 inference graph built in 21.07s!
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
index d790b9588..fbd042eab 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -303,7 +303,7 @@ The compilation steps are:
"target_host parameter is going to be deprecated. "
/workspace/python/tvm/relay/build_module.py:389: DeprecationWarning: Please use input parameter mod (tvm.IRModule) instead of deprecated parameter mod (tvm.relay.function.Function)
DeprecationWarning,
- yolov3-tiny inference graph built in 14.79s!
+ yolov3-tiny inference graph built in 14.74s!
diff --git a/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
index e34b72de0..d1d9add8b 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
@@ -5,7 +5,7 @@
Computation times
=================
-**01:27.666** total execution time for **topic_vta_tutorials_frontend** files:
+**01:27.844** total execution time for **topic_vta_tutorials_frontend** files:
-- **00:46.765**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)
-- **00:40.901**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``)
+- **00:46.696**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)
+- **00:41.147**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``)
diff --git a/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
index 177970eac..912462864 100644
--- a/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
@@ -5,7 +5,7 @@
Computation times
=================
-**00:03.377** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.453** total execution time for **topic_vta_tutorials_optimize** files:
-- **00:02.879**: :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)
-- **00:00.498**: :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``)
+- **00:02.965**: :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)
+- **00:00.488**: :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``)
diff --git a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
index 4bd07ac68..00e60dfe1 100644
--- a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
@@ -5,7 +5,7 @@
Computation times
=================
-**00:00.837** total execution time for **topic_vta_tutorials** files:
+**00:00.839** total execution time for **topic_vta_tutorials** files:
-- **00:00.423**: :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``)
-- **00:00.414**: :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``)
+- **00:00.425**: :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``)
+- **00:00.413**: :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``)
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index 0af62d225..f6d1e291d 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -305,7 +305,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 92.330 ms
+ Execution time of this operator: 93.220 ms
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index 19890930e..71adccb5f 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -280,7 +280,7 @@ standard deviation.
.. code-block:: none
- {'mean': 489.71124089001023, 'median': 489.61681099999623, 'std': 1.282927073204638}
+ {'mean': 489.9442822200035, 'median': 489.9185508000073, 'std': 0.34706001098525147}
@@ -494,29 +494,29 @@ the tuning data to.
.. code-block:: none
-
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 16.84/ 22.92 GFLOPS | Progress: (16/20) | 13.85 s
[Task 1/25] Current/Best: 11.65/ 23.99 GFLOPS | Progress: (20/20) | 16.39 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 13.11/ 21.50 GFLOPS | Progress: (16/20) | 7.75 s
[Task 2/25] Current/Best: 20.34/ 21.50 GFLOPS | Progress: (20/20) | 9.21 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 7.21/ 23.88 GFLOPS | Progress: (16/20) | 10.37 s
[Task 3/25] Current/Best: 12.69/ 23.88 GFLOPS | Progress: (20/20) | 14.81 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 17.43/ 22.53 GFLOPS | Progress: (16/20) | 9.31 s
[Task 4/25] Current/Best: 13.43/ 22.53 GFLOPS | Progress: (20/20) | 11.22 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 11.84/ 23.02 GFLOPS | Progress: (16/20) | 7.85 s
[Task 5/25] Current/Best: 12.12/ 23.02 GFLOPS | Progress: (20/20) | 9.47 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 20.14/ 20.87 GFLOPS | Progress: (16/20) | 8.58 s
[Task 6/25] Current/Best: 3.71/ 20.87 GFLOPS | Progress: (20/20) | 11.03 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 12.28/ 21.21 GFLOPS | Progress: (16/20) | 7.98 s
[Task 7/25] Current/Best: 6.42/ 21.82 GFLOPS | Progress: (20/20) | 10.33 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 18.88/ 18.88 GFLOPS | Progress: (16/20) | 12.65 s
[Task 8/25] Current/Best: 19.68/ 19.68 GFLOPS | Progress: (20/20) | 19.65 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 18.09/ 23.33 GFLOPS | Progress: (16/20) | 14.96 s
[Task 9/25] Current/Best: 9.08/ 23.33 GFLOPS | Progress: (20/20) | 23.41 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 10/25] Current/Best: 19.18/ 20.54 GFLOPS | Progress: (16/20) | 6.14 s
[Task 10/25] Current/Best: 8.85/ 20.54 GFLOPS | Progress: (20/20) | 7.61 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 11.64/ 21.32 GFLOPS | Progress: (16/20) | 8.89 s
[Task 11/25] Current/Best: 19.48/ 21.61 GFLOPS | Progress: (20/20) | 10.89 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 15.46/ 19.05 GFLOPS | Progress: (16/20) | 10.55 s
[Task 12/25] Current/Best: 15.11/ 19.19 GFLOPS | Progress: (20/20) | 12.37 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 12.33/ 21.25 GFLOPS | Progress: (16/20) | 8.95 s
[Task 13/25] Current/Best: 18.80/ 21.25 GFLOPS | Progress: (20/20) | 11.13 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 16.31/ 20.75 GFLOPS | Progress: (16/20) | 8.17 s
[Task 14/25] Current/Best: 17.14/ 20.75 GFLOPS | Progress: (20/20) | 9.55 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 15/25] Current/Best: 20.47/ 22.14 GFLOPS | Progress: (16/20) | 6.35 s
[Task 15/25] Current/Best: 9.68/ 22.14 GFLOPS | Progress: (20/20) | 7.39 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 18.22/ 20.78 GFLOPS | Progress: (16/20) | 6.32 s
[Task 16/25] Current/Best: 10.01/ 22.74 GFLOPS | Progress: (20/20) | 8.38 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 16.54/ 23.47 GFLOPS | Progress: (16/20) | 9.40 s
[Task 17/25] Current/Best: 10.06/ 23.47 GFLOPS | Progress: (20/20) | 11.44 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 10.04/ 19.60 GFLOPS | Progress: (16/20) | 9.58 s
[Task 18/25] Current/Best: 20.80/ 20.80 GFLOPS | Progress: (20/20) | 11.01 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 14.76/ 21.98 GFLOPS | Progress: (16/20) | 11.66 s
[Task 19/25] Current/Best: 2.70/ 23.77 GFLOPS | Progress: (20/20) | 14.37 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 12.42/ 16.82 GFLOPS | Progress: (16/20) | 10.88 s
[Task 20/25] Current/Best: 12.77/ 22.20 GFLOPS | Progress: (20/20) | 12.91 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 21/25] Current/Best: 18.19/ 18.19 GFLOPS | Progress: (16/20) | 7.55 s
[Task 21/25] Current/Best: 4.48/ 18.19 GFLOPS | Progress: (20/20) | 14.80 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 15.55/ 21.68 GFLOPS | Progress: (16/20) | 7.25 s
[Task 22/25] Current/Best: 13.51/ 21.68 GFLOPS | Progress: (20/20) | 8.87 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 6.35/ 21.67 GFLOPS | Progress: (16/20) | 13.51 s
[Task 23/25] Current/Best: 7.93/ 21.67 GFLOPS | Progress: (20/20) | 17.62 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 6.47/ 8.94 GFLOPS | Progress: (16/20) | 13.79 s
[Task 24/25] Current/Best: 3.35/ 8.94 GFLOPS | Progress: (20/20) | 19.67 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 25/25] Current/Best: 5.97/ 8.93 GFLOPS | Progress: (16/20) | 13.49 s
[Task 25/25] Current/Best: 2.89/ 9.62 GFLOPS | Progress: (20/20) | 24.13 s
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 16.75/ 22.83 GFLOPS | Progress: (16/20) | 12.52 s
[Task 1/25] Current/Best: 11.62/ 23.93 GFLOPS | Progress: (20/20) | 15.62 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 12.36/ 21.11 GFLOPS | Progress: (16/20) | 8.73 s
[Task 2/25] Current/Best: 20.40/ 21.11 GFLOPS | Progress: (20/20) | 10.18 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 7.18/ 23.86 GFLOPS | Progress: (16/20) | 10.43 s
[Task 3/25] Current/Best: 12.74/ 23.86 GFLOPS | Progress: (20/20) | 14.86 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 17.42/ 21.93 GFLOPS | Progress: (16/20) | 9.24 s
[Task 4/25] Current/Best: 13.81/ 21.93 GFLOPS | Progress: (20/20) | 11.14 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 11.96/ 23.10 GFLOPS | Progress: (16/20) | 7.91 s
[Task 5/25] Current/Best: 12.13/ 23.10 GFLOPS | Progress: (20/20) | 9.54 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 20.00/ 20.94 GFLOPS | Progress: (16/20) | 8.58 s
[Task 6/25] Current/Best: 3.75/ 20.94 GFLOPS | Progress: (20/20) | 11.03 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 12.27/ 21.21 GFLOPS | Progress: (16/20) | 8.08 s
[Task 7/25] Current/Best: 6.39/ 21.78 GFLOPS | Progress: (20/20) | 10.44 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 19.03/ 19.03 GFLOPS | Progress: (16/20) | 12.77 s
[Task 8/25] Current/Best: 20.11/ 20.11 GFLOPS | Progress: (20/20) | 19.78 s Done.
+
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 17.82/ 23.32 GFLOPS | Progress: (16/20) | 15.05 s
[Task 9/25] Current/Best: 8.95/ 23.32 GFLOPS | Progress: (20/20) | 23.70 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 10/25] Current/Best: 19.25/ 20.66 GFLOPS | Progress: (16/20) | 6.16 s
[Task 10/25] Current/Best: 8.85/ 20.66 GFLOPS | Progress: (20/20) | 7.66 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 13.59/ 21.26 GFLOPS | Progress: (16/20) | 8.91 s
[Task 11/25] Current/Best: 19.51/ 21.26 GFLOPS | Progress: (20/20) | 10.90 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 15.53/ 19.18 GFLOPS | Progress: (16/20) | 10.36 s
[Task 12/25] Current/Best: 15.22/ 19.18 GFLOPS | Progress: (20/20) | 12.19 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 12.32/ 21.70 GFLOPS | Progress: (16/20) | 8.92 s
[Task 13/25] Current/Best: 18.98/ 21.70 GFLOPS | Progress: (20/20) | 11.13 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 16.01/ 20.30 GFLOPS | Progress: (16/20) | 8.07 s
[Task 14/25] Current/Best: 17.24/ 20.30 GFLOPS | Progress: (20/20) | 9.44 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 15/25] Current/Best: 20.47/ 22.39 GFLOPS | Progress: (16/20) | 6.51 s
[Task 15/25] Current/Best: 9.71/ 22.39 GFLOPS | Progress: (20/20) | 7.63 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 17.87/ 20.89 GFLOPS | Progress: (16/20) | 6.30 s
[Task 16/25] Current/Best: 9.96/ 22.06 GFLOPS | Progress: (20/20) | 8.39 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 16.52/ 23.40 GFLOPS | Progress: (16/20) | 9.40 s
[Task 17/25] Current/Best: 10.05/ 23.40 GFLOPS | Progress: (20/20) | 11.46 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 10.03/ 18.88 GFLOPS | Progress: (16/20) | 9.66 s
[Task 18/25] Current/Best: 20.88/ 20.88 GFLOPS | Progress: (20/20) | 11.11 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 13.68/ 21.77 GFLOPS | Progress: (16/20) | 11.57 s
[Task 19/25] Current/Best: 2.70/ 23.69 GFLOPS | Progress: (20/20) | 14.24 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 12.54/ 16.84 GFLOPS | Progress: (16/20) | 10.97 s
[Task 20/25] Current/Best: 13.67/ 22.19 GFLOPS | Progress: (20/20) | 13.00 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 21/25] Current/Best: 18.17/ 18.17 GFLOPS | Progress: (16/20) | 7.57 s
[Task 21/25] Current/Best: 4.47/ 18.17 GFLOPS | Progress: (20/20) | 14.70 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 15.60/ 22.08 GFLOPS | Progress: (16/20) | 7.13 s
[Task 22/25] Current/Best: 14.55/ 22.08 GFLOPS | Progress: (20/20) | 8.72 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 6.35/ 21.90 GFLOPS | Progress: (16/20) | 13.48 s
[Task 23/25] Current/Best: 7.95/ 21.90 GFLOPS | Progress: (20/20) | 17.58 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 6.74/ 9.04 GFLOPS | Progress: (16/20) | 13.77 s
[Task 24/25] Current/Best: 3.42/ 9.04 GFLOPS | Progress: (20/20) | 19.75 s Done.
+
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 5.86/ 9.61 GFLOPS | Progress: (16/20) | 13.50 s
[Task 25/25] Current/Best: 2.92/ 9.67 GFLOPS | Progress: (20/20) | 24.17 s
The output from this tuning process will look something like this:
@@ -658,8 +658,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 411.1099095999998, 'median': 408.52510024997173, 'std': 5.92846754026032}
- unoptimized: {'mean': 489.71124089001023, 'median': 489.61681099999623, 'std': 1.282927073204638}
+ optimized: {'mean': 411.8927176699981, 'median': 405.329046550014, 'std': 9.447091200706092}
+ unoptimized: {'mean': 489.9442822200035, 'median': 489.9185508000073, 'std': 0.34706001098525147}
@@ -679,7 +679,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 8 minutes 33.498 seconds)
+ **Total running time of the script:** ( 8 minutes 37.531 seconds)
.. _sphx_glr_download_tutorial_autotvm_relay_x86.py:
diff --git a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
index 45722409d..383ac658b 100644
--- a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
+++ b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
@@ -235,7 +235,7 @@ device and returns the measured cost. Network overhead is excluded.
.. code-block:: none
- 1.295e-07 secs/op
+ 1.336e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index 34825481b..906b267b6 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -233,7 +233,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x2211ba60)), stage(b, placeholder(b, 0x21296860)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(mi [...]
+ [stage(a, placeholder(a, 0x20941770)), stage(b, placeholder(b, 0xc215c90)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min [...]
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index ab13889f2..5f4a34f49 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,17 +5,17 @@
Computation times
=================
-**10:58.179** total execution time for **tutorial** files:
+**11:15.463** total execution time for **tutorial** files:
-- **08:33.498**: :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)
-- **01:00.727**: :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)
-- **00:32.810**: :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``)
-- **00:25.709**: :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)
-- **00:23.918**: :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)
-- **00:00.693**: :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)
-- **00:00.517**: :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)
-- **00:00.174**: :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``)
-- **00:00.038**: :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)
-- **00:00.036**: :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``)
+- **08:37.531**: :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)
+- **00:58.614**: :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)
+- **00:46.875**: :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``)
+- **00:26.535**: :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)
+- **00:24.288**: :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)
+- **00:00.715**: :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)
+- **00:00.574**: :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)
+- **00:00.206**: :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``)
+- **00:00.034**: :ref:`sphx_glr_tutorial_install.py` (``install.py``)
+- **00:00.032**: :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)
- **00:00.031**: :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``)
-- **00:00.029**: :ref:`sphx_glr_tutorial_install.py` (``install.py``)
+- **00:00.030**: :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``)
diff --git a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
index 642bdb546..98e559c58 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -253,7 +253,7 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
Numpy running time: 0.000008
- naive: 0.000008
+ naive: 0.000007
@@ -397,7 +397,7 @@ factor to be the number of threads on your CPU.
.. code-block:: none
- vector: 0.000024
+ vector: 0.000025
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [(stride: int32*n: int32)], [], type="auto"),
@@ -447,10 +447,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 8.20971999928588e-06 1.0
- naive 7.771e-06 0.9465609059353983
- parallel 7.0948e-06 0.8641951248784537
- vector 2.44877e-05 2.982769205542948
+ numpy 7.702709999648506e-06 1.0
+ naive 6.6650000000000006e-06 0.8652798820550354
+ parallel 6.9181e-06 0.8981384474185956
+ vector 2.45071e-05 3.181620494750331
@@ -839,7 +839,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.017535
+ Numpy running time: 0.018414
@@ -897,7 +897,7 @@ optimizations.
/workspace/python/tvm/driver/build_module.py:264: UserWarning: target_host parameter is going to be deprecated. Please pass in tvm.target.Target(target, host=target_host) instead.
"target_host parameter is going to be deprecated. "
- none: 3.426943
+ none: 3.254004
@@ -996,7 +996,7 @@ schedule.
.. code-block:: none
- blocking: 0.294216
+ blocking: 0.295370
@@ -1088,7 +1088,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.331346
+ vectorization: 0.330980
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [1048576], []),
@@ -1160,7 +1160,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.112761
+ loop permutation: 0.113367
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [1048576], []),
@@ -1257,7 +1257,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.108503
+ array packing: 0.108772
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [1048576], []),
@@ -1348,7 +1348,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.109923
+ block caching: 0.109201
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [1048576], []),
@@ -1432,7 +1432,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.144371
+ parallelization: 0.143560
@main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
buffers = {A: Buffer(A_2: Pointer(float32), float32, [1048576], []),
@@ -1511,13 +1511,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.4269425639 1.0
- blocking 0.29421639650000003 0.08585390359305288
- vectorization 0.3313455339 0.09668838263893029
- loop permutation 0.1127613861 0.03290437000253455
- array packing 0.1085034061 0.031661868874895506
- block caching 0.1099231018 0.03207614360332408
- parallelization 0.1443710097 0.042128225672886654
+ none 3.2540043058 1.0
+ blocking 0.2953703983 0.09077136061975274
+ vectorization 0.3309797321 0.10171459561686975
+ loop permutation 0.11336696629999998 0.03483921828189733
+ array packing 0.10877214449999999 0.033427166739184215
+ block caching 0.1092009911 0.0335589571609841
+ parallelization 0.1435596609 0.044117846016404
@@ -1552,11 +1552,6 @@ operations with tunable parameters that allows you to automatically optimize
the computation for specific platforms.
-.. rst-class:: sphx-glr-timing
-
- **Total running time of the script:** ( 1 minutes 0.727 seconds)
-
-
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
diff --git a/docs/commit_hash b/docs/commit_hash
index 613a09e71..7a2020d2c 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-a71536a130685a50582eea8c993030872cddb145
+0cd4dd2f2d6cab265844de0cb8745e0de8d22571
diff --git a/docs/how_to/compile_models/from_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index 701b6f61e..ffca0f0a0 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -401,7 +401,7 @@
</div>
<img alt="../../_images/sphx_glr_from_mxnet_001.png" class="sphx-glr-single-img" src="../../_images/sphx_glr_from_mxnet_001.png" />
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipe4724699-b020-4e8b-9f7b-3ca7e7462716 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip1f7a3cf1-7d31-4f66-a000-26fa89c1904f from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
x (1, 3, 224, 224)
</pre></div>
</div>
diff --git a/docs/how_to/compile_models/from_oneflow.html b/docs/how_to/compile_models/from_oneflow.html
index 735bcc2d3..949cfd1ab 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -406,98 +406,41 @@ python3 -m pip install -f https://release.oneflow.info <span class="nv">oneflow<
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: "https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip" to /workspace/.oneflow/flowvision_cache/resnet18.zip
0%| | 0.00/41.5M [00:00<?, ?B/s]
- 0%| | 16.0k/41.5M [00:00<07:34, 95.6kB/s]
- 0%| | 48.0k/41.5M [00:00<04:47, 151kB/s]
- 0%| | 104k/41.5M [00:00<03:05, 234kB/s]
- 0%| | 160k/41.5M [00:00<02:38, 273kB/s]
- 1%| | 224k/41.5M [00:00<02:18, 312kB/s]
- 1%| | 288k/41.5M [00:01<02:08, 335kB/s]
- 1%| | 360k/41.5M [00:01<01:58, 365kB/s]
- 1%|1 | 432k/41.5M [00:01<01:51, 385kB/s]
- 1%|1 | 512k/41.5M [00:01<01:43, 413kB/s]
- 1%|1 | 592k/41.5M [00:01<01:39, 433kB/s]
- 2%|1 | 680k/41.5M [00:01<01:33, 460kB/s]
- 2%|1 | 768k/41.5M [00:02<01:29, 479kB/s]
- 2%|2 | 864k/41.5M [00:02<01:24, 507kB/s]
- 2%|2 | 960k/41.5M [00:02<01:20, 526kB/s]
- 3%|2 | 1.04M/41.5M [00:02<01:16, 554kB/s]
- 3%|2 | 1.15M/41.5M [00:02<01:12, 587kB/s]
- 3%|3 | 1.26M/41.5M [00:02<01:09, 611kB/s]
- 3%|3 | 1.38M/41.5M [00:03<01:05, 642kB/s]
- 4%|3 | 1.49M/41.5M [00:03<01:03, 663kB/s]
- 4%|3 | 1.62M/41.5M [00:03<01:00, 692kB/s]
- 4%|4 | 1.75M/41.5M [00:03<00:57, 727kB/s]
- 5%|4 | 1.88M/41.5M [00:03<00:55, 751kB/s]
- 5%|4 | 2.02M/41.5M [00:03<00:52, 783kB/s]
- 5%|5 | 2.17M/41.5M [00:04<00:50, 819kB/s]
- 6%|5 | 2.33M/41.5M [00:04<00:47, 858kB/s]
- 6%|6 | 2.49M/41.5M [00:04<00:45, 900kB/s]
- 6%|6 | 2.66M/41.5M [00:04<00:43, 944kB/s]
- 7%|6 | 2.84M/41.5M [00:04<00:41, 988kB/s]
- 7%|7 | 3.03M/41.5M [00:04<00:39, 1.03MB/s]
- 8%|7 | 3.23M/41.5M [00:05<00:36, 1.09MB/s]
- 8%|8 | 3.44M/41.5M [00:05<00:35, 1.14MB/s]
- 9%|8 | 3.66M/41.5M [00:05<00:33, 1.19MB/s]
- 9%|9 | 3.88M/41.5M [00:05<00:31, 1.25MB/s]
- 10%|9 | 4.12M/41.5M [00:05<00:29, 1.32MB/s]
- 11%|# | 4.38M/41.5M [00:06<00:28, 1.38MB/s]
- 11%|#1 | 4.63M/41.5M [00:06<00:26, 1.44MB/s]
- 12%|#1 | 4.91M/41.5M [00:06<00:25, 1.50MB/s]
- 13%|#2 | 5.19M/41.5M [00:06<00:24, 1.57MB/s]
- 13%|#3 | 5.48M/41.5M [00:06<00:23, 1.64MB/s]
- 14%|#3 | 5.80M/41.5M [00:06<00:21, 1.72MB/s]
- 15%|#4 | 6.13M/41.5M [00:07<00:20, 1.81MB/s]
- 16%|#5 | 6.48M/41.5M [00:07<00:19, 1.90MB/s]
- 16%|#6 | 6.84M/41.5M [00:07<00:18, 1.98MB/s]
- 17%|#7 | 7.22M/41.5M [00:07<00:17, 2.09MB/s]
- 18%|#8 | 7.62M/41.5M [00:07<00:16, 2.19MB/s]
- 19%|#9 | 8.03M/41.5M [00:07<00:14, 2.49MB/s]
- 20%|## | 8.47M/41.5M [00:08<00:12, 2.75MB/s]
- 21%|##1 | 8.75M/41.5M [00:08<00:13, 2.62MB/s]
- 22%|##1 | 9.01M/41.5M [00:08<00:13, 2.46MB/s]
- 23%|##2 | 9.41M/41.5M [00:08<00:12, 2.71MB/s]
- 24%|##3 | 9.91M/41.5M [00:08<00:11, 2.82MB/s]
- 25%|##5 | 10.4M/41.5M [00:08<00:10, 3.20MB/s]
- 26%|##5 | 10.8M/41.5M [00:08<00:10, 3.07MB/s]
- 27%|##6 | 11.1M/41.5M [00:08<00:11, 2.86MB/s]
- 28%|##7 | 11.5M/41.5M [00:09<00:10, 2.90MB/s]
- 29%|##9 | 12.1M/41.5M [00:09<00:09, 3.11MB/s]
- 31%|### | 12.7M/41.5M [00:09<00:09, 3.29MB/s]
- 32%|###2 | 13.4M/41.5M [00:09<00:08, 3.46MB/s]
- 34%|###3 | 14.0M/41.5M [00:09<00:08, 3.58MB/s]
- 35%|###5 | 14.6M/41.5M [00:09<00:06, 4.04MB/s]
- 37%|###6 | 15.3M/41.5M [00:10<00:05, 4.58MB/s]
- 38%|###8 | 15.8M/41.5M [00:10<00:06, 4.20MB/s]
- 39%|###9 | 16.2M/41.5M [00:10<00:06, 3.94MB/s]
- 41%|#### | 16.8M/41.5M [00:10<00:06, 3.84MB/s]
- 42%|####2 | 17.6M/41.5M [00:10<00:06, 4.11MB/s]
- 44%|####4 | 18.4M/41.5M [00:10<00:05, 4.38MB/s]
- 46%|####6 | 19.2M/41.5M [00:10<00:04, 4.90MB/s]
- 48%|####8 | 20.1M/41.5M [00:11<00:04, 5.58MB/s]
- 50%|####9 | 20.6M/41.5M [00:11<00:04, 5.29MB/s]
- 51%|##### | 21.2M/41.5M [00:11<00:04, 4.94MB/s]
- 53%|#####3 | 22.0M/41.5M [00:11<00:03, 5.37MB/s]
- 55%|#####5 | 23.0M/41.5M [00:11<00:03, 6.18MB/s]
- 57%|#####6 | 23.6M/41.5M [00:11<00:03, 6.09MB/s]
- 58%|#####8 | 24.2M/41.5M [00:11<00:03, 5.57MB/s]
- 61%|###### | 25.1M/41.5M [00:11<00:02, 6.61MB/s]
- 62%|######2 | 25.8M/41.5M [00:12<00:02, 6.50MB/s]
- 64%|######3 | 26.5M/41.5M [00:12<00:02, 6.04MB/s]
- 66%|######6 | 27.5M/41.5M [00:12<00:02, 7.27MB/s]
- 68%|######7 | 28.2M/41.5M [00:12<00:01, 7.13MB/s]
- 70%|######9 | 28.9M/41.5M [00:12<00:01, 6.61MB/s]
- 72%|#######2 | 30.1M/41.5M [00:12<00:01, 8.01MB/s]
- 74%|#######4 | 30.9M/41.5M [00:12<00:01, 7.87MB/s]
- 76%|#######6 | 31.6M/41.5M [00:12<00:01, 7.29MB/s]
- 79%|#######9 | 32.9M/41.5M [00:13<00:01, 8.81MB/s]
- 81%|########1 | 33.8M/41.5M [00:13<00:00, 8.65MB/s]
- 83%|########3 | 34.6M/41.5M [00:13<00:00, 8.01MB/s]
- 86%|########6 | 35.9M/41.5M [00:13<00:00, 8.85MB/s]
- 90%|########9 | 37.3M/41.5M [00:13<00:00, 10.3MB/s]
- 92%|#########2| 38.3M/41.5M [00:13<00:00, 8.94MB/s]
- 94%|#########4| 39.2M/41.5M [00:13<00:00, 8.37MB/s]
- 97%|#########7| 40.3M/41.5M [00:13<00:00, 8.72MB/s]
-100%|##########| 41.5M/41.5M [00:14<00:00, 3.11MB/s]
+ 0%| | 16.0k/41.5M [00:00<07:40, 94.3kB/s]
+ 0%| | 48.0k/41.5M [00:00<04:51, 149kB/s]
+ 0%| | 104k/41.5M [00:00<03:07, 231kB/s]
+ 0%| | 208k/41.5M [00:00<01:53, 381kB/s]
+ 1%| | 424k/41.5M [00:00<01:01, 700kB/s]
+ 2%|2 | 864k/41.5M [00:01<00:31, 1.34MB/s]
+ 4%|4 | 1.69M/41.5M [00:01<00:16, 2.56MB/s]
+ 8%|7 | 3.16M/41.5M [00:01<00:08, 4.55MB/s]
+ 11%|#1 | 4.62M/41.5M [00:01<00:06, 5.88MB/s]
+ 15%|#4 | 6.09M/41.5M [00:01<00:05, 6.79MB/s]
+ 18%|#8 | 7.56M/41.5M [00:01<00:04, 7.41MB/s]
+ 22%|##1 | 9.04M/41.5M [00:02<00:04, 7.85MB/s]
+ 25%|##5 | 10.5M/41.5M [00:02<00:03, 8.13MB/s]
+ 29%|##8 | 12.0M/41.5M [00:02<00:03, 8.34MB/s]
+ 32%|###2 | 13.4M/41.5M [00:02<00:03, 8.48MB/s]
+ 36%|###5 | 14.9M/41.5M [00:02<00:03, 8.58MB/s]
+ 39%|###9 | 16.4M/41.5M [00:02<00:03, 8.64MB/s]
+ 43%|####3 | 17.8M/41.5M [00:03<00:02, 8.69MB/s]
+ 47%|####6 | 19.3M/41.5M [00:03<00:02, 8.73MB/s]
+ 50%|##### | 20.8M/41.5M [00:03<00:02, 8.74MB/s]
+ 54%|#####3 | 22.2M/41.5M [00:03<00:02, 8.76MB/s]
+ 57%|#####7 | 23.7M/41.5M [00:03<00:02, 8.78MB/s]
+ 61%|###### | 25.2M/41.5M [00:04<00:01, 8.79MB/s]
+ 64%|######4 | 26.6M/41.5M [00:04<00:01, 8.79MB/s]
+ 68%|######7 | 28.1M/41.5M [00:04<00:01, 8.79MB/s]
+ 71%|#######1 | 29.6M/41.5M [00:04<00:01, 8.79MB/s]
+ 75%|#######4 | 31.0M/41.5M [00:04<00:01, 8.80MB/s]
+ 78%|#######8 | 32.5M/41.5M [00:04<00:01, 8.80MB/s]
+ 82%|########1 | 34.0M/41.5M [00:05<00:00, 8.81MB/s]
+ 85%|########5 | 35.5M/41.5M [00:05<00:00, 8.81MB/s]
+ 89%|########8 | 36.9M/41.5M [00:05<00:00, 8.80MB/s]
+ 93%|#########2| 38.4M/41.5M [00:05<00:00, 9.84MB/s]
+ 96%|#########6| 39.8M/41.5M [00:05<00:00, 10.3MB/s]
+ 98%|#########8| 40.9M/41.5M [00:05<00:00, 10.1MB/s]
+100%|##########| 41.5M/41.5M [00:05<00:00, 7.32MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_paddle.html b/docs/how_to/compile_models/from_paddle.html
index 30b4942fa..2cb9f52e0 100644
--- a/docs/how_to/compile_models/from_paddle.html
+++ b/docs/how_to/compile_models/from_paddle.html
@@ -469,7 +469,7 @@ A quick solution is</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>TVM prediction top-1 id: 282, class name: 282: 'tiger cat',
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 10.497 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 6.877 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-paddle-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/16269b77359771348d507395692524cf/from_paddle.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">from_paddle.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index 06cd1868b..c6f852066 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -387,8 +387,8 @@ be unstable.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
0%| | 0.00/44.7M [00:00<?, ?B/s]
- 43%|####3 | 19.4M/44.7M [00:00<00:00, 203MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 238MB/s]
+ 44%|####3 | 19.6M/44.7M [00:00<00:00, 206MB/s]
+100%|##########| 44.7M/44.7M [00:00<00:00, 239MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/sg_execution_times.html b/docs/how_to/compile_models/sg_execution_times.html
index fd5c202a3..67fedf73c 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -300,18 +300,18 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-compile-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>05:24.411</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>05:31.303</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<ul class="simple">
-<li><p><strong>01:10.497</strong>: <a class="reference internal" href="from_paddle.html#sphx-glr-how-to-compile-models-from-paddle-py"><span class="std std-ref">Compile PaddlePaddle Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_paddle.py</span></code>)</p></li>
-<li><p><strong>00:59.180</strong>: <a class="reference internal" href="from_tensorflow.html#sphx-glr-how-to-compile-models-from-tensorflow-py"><span class="std std-ref">Compile Tensorflow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tensorflow.py</span></code>)</p></li>
-<li><p><strong>00:56.550</strong>: <a class="reference internal" href="from_darknet.html#sphx-glr-how-to-compile-models-from-darknet-py"><span class="std std-ref">Compile YOLO-V2 and YOLO-V3 in DarkNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_darknet.py</span></code>)</p></li>
-<li><p><strong>00:37.552</strong>: <a class="reference internal" href="from_oneflow.html#sphx-glr-how-to-compile-models-from-oneflow-py"><span class="std std-ref">Compile OneFlow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_oneflow.py</span></code>)</p></li>
-<li><p><strong>00:24.066</strong>: <a class="reference internal" href="from_tflite.html#sphx-glr-how-to-compile-models-from-tflite-py"><span class="std std-ref">Compile TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tflite.py</span></code>)</p></li>
-<li><p><strong>00:21.036</strong>: <a class="reference internal" href="from_mxnet.html#sphx-glr-how-to-compile-models-from-mxnet-py"><span class="std std-ref">Compile MXNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_mxnet.py</span></code>)</p></li>
-<li><p><strong>00:20.886</strong>: <a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></li>
-<li><p><strong>00:18.870</strong>: <a class="reference internal" href="from_pytorch.html#sphx-glr-how-to-compile-models-from-pytorch-py"><span class="std std-ref">Compile PyTorch Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_pytorch.py</span></code>)</p></li>
-<li><p><strong>00:13.327</strong>: <a class="reference internal" href="from_keras.html#sphx-glr-how-to-compile-models-from-keras-py"><span class="std std-ref">Compile Keras Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_keras.py</span></code>)</p></li>
-<li><p><strong>00:02.447</strong>: <a class="reference internal" href="from_onnx.html#sphx-glr-how-to-compile-models-from-onnx-py"><span class="std std-ref">Compile ONNX Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_onnx.py</span></code>)</p></li>
+<li><p><strong>01:06.877</strong>: <a class="reference internal" href="from_paddle.html#sphx-glr-how-to-compile-models-from-paddle-py"><span class="std std-ref">Compile PaddlePaddle Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_paddle.py</span></code>)</p></li>
+<li><p><strong>00:59.147</strong>: <a class="reference internal" href="from_tensorflow.html#sphx-glr-how-to-compile-models-from-tensorflow-py"><span class="std std-ref">Compile Tensorflow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tensorflow.py</span></code>)</p></li>
+<li><p><strong>00:57.219</strong>: <a class="reference internal" href="from_darknet.html#sphx-glr-how-to-compile-models-from-darknet-py"><span class="std std-ref">Compile YOLO-V2 and YOLO-V3 in DarkNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_darknet.py</span></code>)</p></li>
+<li><p><strong>00:39.405</strong>: <a class="reference internal" href="from_tflite.html#sphx-glr-how-to-compile-models-from-tflite-py"><span class="std std-ref">Compile TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tflite.py</span></code>)</p></li>
+<li><p><strong>00:30.243</strong>: <a class="reference internal" href="from_oneflow.html#sphx-glr-how-to-compile-models-from-oneflow-py"><span class="std std-ref">Compile OneFlow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_oneflow.py</span></code>)</p></li>
+<li><p><strong>00:21.431</strong>: <a class="reference internal" href="from_mxnet.html#sphx-glr-how-to-compile-models-from-mxnet-py"><span class="std std-ref">Compile MXNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_mxnet.py</span></code>)</p></li>
+<li><p><strong>00:20.708</strong>: <a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></li>
+<li><p><strong>00:19.861</strong>: <a class="reference internal" href="from_pytorch.html#sphx-glr-how-to-compile-models-from-pytorch-py"><span class="std std-ref">Compile PyTorch Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_pytorch.py</span></code>)</p></li>
+<li><p><strong>00:13.896</strong>: <a class="reference internal" href="from_keras.html#sphx-glr-how-to-compile-models-from-keras-py"><span class="std std-ref">Compile Keras Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_keras.py</span></code>)</p></li>
+<li><p><strong>00:02.516</strong>: <a class="reference internal" href="from_onnx.html#sphx-glr-how-to-compile-models-from-onnx-py"><span class="std std-ref">Compile ONNX Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_onnx.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/deploy_models/deploy_model_on_android.html b/docs/how_to/deploy_models/deploy_model_on_android.html
index 015799d23..316a5e13e 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -627,7 +627,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.4666 15.4579 15.6077 15.4158 0.0530
+ 16.0870 16.0837 16.1489 16.0308 0.0370
</pre></div>
</div>
</div>
diff --git a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
index 80305ee60..309bf49f5 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -409,13 +409,14 @@ be unstable.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: "https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth" to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
0%| | 0.00/170M [00:00<?, ?B/s]
- 13%|#3 | 22.2M/170M [00:00<00:00, 233MB/s]
- 30%|##9 | 50.4M/170M [00:00<00:00, 270MB/s]
- 45%|####4 | 76.1M/170M [00:00<00:00, 255MB/s]
- 60%|#####9 | 101M/170M [00:00<00:00, 259MB/s]
- 76%|#######6 | 130M/170M [00:00<00:00, 272MB/s]
- 93%|#########2| 158M/170M [00:00<00:00, 280MB/s]
-100%|##########| 170M/170M [00:00<00:00, 272MB/s]
+ 3%|2 | 4.44M/170M [00:00<00:03, 46.4MB/s]
+ 5%|5 | 8.87M/170M [00:00<00:03, 45.5MB/s]
+ 22%|##1 | 36.5M/170M [00:00<00:00, 156MB/s]
+ 38%|###8 | 65.0M/170M [00:00<00:00, 212MB/s]
+ 55%|#####5 | 93.4M/170M [00:00<00:00, 243MB/s]
+ 72%|#######1 | 122M/170M [00:00<00:00, 262MB/s]
+ 89%|########8 | 150M/170M [00:00<00:00, 273MB/s]
+100%|##########| 170M/170M [00:00<00:00, 231MB/s]
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:3878: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
for i in range(dim)
/usr/local/lib/python3.7/dist-packages/torchvision/models/detection/anchor_utils.py:127: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
@@ -513,7 +514,7 @@ torchvision rcnn models.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Get 9 valid boxes
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 56.734 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 1.543 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-object-detection-pytorch-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/7795da4b258c8feff986668b95ef57ad/deploy_object_detection_pytorch.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_object_detection_pytorch.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_prequantized.html b/docs/how_to/deploy_models/deploy_prequantized.html
index 441eccaee..4d05ee1b2 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -450,7 +450,7 @@ training. Other models require a full post training calibration.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: "https://download.pytorch.org/models/mobilenet_v2-b0353104.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
0%| | 0.00/13.6M [00:00<?, ?B/s]
-100%|##########| 13.6M/13.6M [00:00<00:00, 182MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 185MB/s]
</pre></div>
</div>
</div>
@@ -544,7 +544,7 @@ output values are identical out of 1000 outputs from mobilenet v2.</p>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 90.5653 90.6645 91.4601 90.0798 0.2444
+ 90.3016 90.2044 90.8585 89.8834 0.2526
</pre></div>
</div>
<div class="admonition note">
@@ -583,7 +583,7 @@ This includes support for the VNNI 8 bit dot product instruction (CascadeLake or
<div class="section" id="deploy-a-quantized-tflite-model">
<h2>Deploy a quantized TFLite Model<a class="headerlink" href="#deploy-a-quantized-tflite-model" title="Permalink to this headline">¶</a></h2>
<p>TODO</p>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 3.979 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 4.650 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/fb8217c13f4351224c6cf3aacf1a87fc/deploy_prequantized.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_prequantized.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_prequantized_tflite.html b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
index 4bb69e90f..f8b77e7ce 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -545,7 +545,7 @@ TFLite Top-5 labels: [387 102 386 341 349]
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 116.9583 116.7426 119.2221 115.7829 0.8334
+ 120.5670 120.5832 121.4224 119.8656 0.2797
</pre></div>
</div>
<div class="admonition note">
@@ -573,7 +573,7 @@ network for ARM CPU</span></a>.</p></li>
</ul>
</div></blockquote>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 56.558 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 56.617 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-tflite-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/56691c7a27d45da61d112276334640d3/deploy_prequantized_tflite.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_prequantized_tflite.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_quantized.html b/docs/how_to/deploy_models/deploy_quantized.html
index bab69bc6a..852895a41 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -482,7 +482,7 @@ for calibration. But the accuracy might be impacted.</p>
DeprecationWarning,
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 11.625 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 14.102 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-quantized-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/7810ecf51bfc05f7d5e8a400ac3e815d/deploy_quantized.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_quantized.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
index 25fde34ee..1d215a1de 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -415,22 +415,22 @@ to your device.</p>
Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
0%| | 0/132723 [00:00<?, ?KB/s]
- 5%|5 | 7046/132723 [00:00<00:01, 70452.51KB/s]
- 12%|#1 | 15578/132723 [00:00<00:01, 79189.87KB/s]
- 18%|#8 | 24179/132723 [00:00<00:01, 82297.79KB/s]
- 25%|##4 | 32822/132723 [00:00<00:01, 83926.10KB/s]
- 31%|###1 | 41433/132723 [00:00<00:01, 84712.06KB/s]
- 38%|###7 | 50032/132723 [00:00<00:00, 85139.87KB/s]
- 44%|####4 | 58637/132723 [00:00<00:00, 85433.22KB/s]
- 51%|##### | 67181/132723 [00:00<00:00, 85428.20KB/s]
- 57%|#####7 | 75791/132723 [00:00<00:00, 85635.41KB/s]
- 64%|######3 | 84355/132723 [00:01<00:00, 85524.31KB/s]
- 70%|####### | 92908/132723 [00:01<00:00, 85273.26KB/s]
- 76%|#######6 | 101436/132723 [00:01<00:00, 85171.71KB/s]
- 83%|########2 | 109954/132723 [00:01<00:00, 85144.31KB/s]
- 89%|########9 | 118469/132723 [00:01<00:00, 84959.97KB/s]
- 96%|#########5| 126975/132723 [00:01<00:00, 84989.14KB/s]
-100%|##########| 132723/132723 [00:01<00:00, 84426.01KB/s]
+ 5%|4 | 6422/132723 [00:00<00:01, 64211.95KB/s]
+ 11%|#1 | 15033/132723 [00:00<00:01, 77090.48KB/s]
+ 18%|#7 | 23723/132723 [00:00<00:01, 81565.56KB/s]
+ 24%|##4 | 32392/132723 [00:00<00:01, 83583.28KB/s]
+ 31%|### | 41113/132723 [00:00<00:01, 84880.23KB/s]
+ 38%|###7 | 49818/132723 [00:00<00:00, 85616.43KB/s]
+ 44%|####4 | 58477/132723 [00:00<00:00, 85932.03KB/s]
+ 51%|##### | 67168/132723 [00:00<00:00, 86238.65KB/s]
+ 57%|#####7 | 75937/132723 [00:00<00:00, 86688.92KB/s]
+ 64%|######3 | 84687/132723 [00:01<00:00, 86937.59KB/s]
+ 70%|####### | 93422/132723 [00:01<00:00, 87061.35KB/s]
+ 77%|#######6 | 102196/132723 [00:01<00:00, 87265.66KB/s]
+ 84%|########3 | 110927/132723 [00:01<00:00, 87276.34KB/s]
+ 90%|######### | 119696/132723 [00:01<00:00, 87399.54KB/s]
+ 97%|#########6| 128436/132723 [00:01<00:00, 87391.35KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 85534.95KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -475,7 +475,7 @@ Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from h
</pre></div>
</div>
<img alt="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" class="sphx-glr-single-img" src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" />
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 35.845 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 37.568 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-ssd-gluoncv-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/cccb17d28e5e8b2e94ea8cd5ec59f6ed/deploy_ssd_gluoncv.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_ssd_gluoncv.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/sg_execution_times.html b/docs/how_to/deploy_models/sg_execution_times.html
index 67eaa845e..312055278 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -300,16 +300,16 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-deploy-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>10:34.413</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>11:44.196</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<ul class="simple">
-<li><p><strong>02:56.734</strong>: <a class="reference internal" href="deploy_object_detection_pytorch.html#sphx-glr-how-to-deploy-models-deploy-object-detection-pytorch-py"><span class="std std-ref">Compile PyTorch Object Detection Models</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_object_detection_pytorch.py</span></code>)</p></li>
-<li><p><strong>02:35.845</strong>: <a class="reference internal" href="deploy_ssd_gluoncv.html#sphx-glr-how-to-deploy-models-deploy-ssd-gluoncv-py"><span class="std std-ref">Deploy Single Shot Multibox Detector(SSD) model</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_ssd_gluoncv.py</span></code>)</p></li>
-<li><p><strong>01:56.558</strong>: <a class="reference internal" href="deploy_prequantized_tflite.html#sphx-glr-how-to-deploy-models-deploy-prequantized-tflite-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM - Part 3 (TFLite)</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized_tflite.py</span></code>)</p></li>
-<li><p><strong>01:11.625</strong>: <a class="reference internal" href="deploy_quantized.html#sphx-glr-how-to-deploy-models-deploy-quantized-py"><span class="std std-ref">Deploy a Quantized Model on Cuda</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_quantized.py</span></code>)</p></li>
-<li><p><strong>01:03.979</strong>: <a class="reference internal" href="deploy_prequantized.html#sphx-glr-how-to-deploy-models-deploy-prequantized-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized.py</span></code>)</p></li>
-<li><p><strong>00:27.872</strong>: <a class="reference internal" href="deploy_model_on_android.html#sphx-glr-how-to-deploy-models-deploy-model-on-android-py"><span class="std std-ref">Deploy the Pretrained Model on Android</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_android.py</span></code>)</p></li>
-<li><p><strong>00:21.621</strong>: <a class="reference internal" href="deploy_model_on_rasp.html#sphx-glr-how-to-deploy-models-deploy-model-on-rasp-py"><span class="std std-ref">Deploy the Pretrained Model on Raspberry Pi</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_rasp.py</span></code>)</p></li>
-<li><p><strong>00:00.177</strong>: <a class="reference internal" href="deploy_sparse.html#sphx-glr-how-to-deploy-models-deploy-sparse-py"><span class="std std-ref">Deploy a Hugging Face Pruned Model on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_sparse.py</span></code>)</p></li>
+<li><p><strong>03:01.543</strong>: <a class="reference internal" href="deploy_object_detection_pytorch.html#sphx-glr-how-to-deploy-models-deploy-object-detection-pytorch-py"><span class="std std-ref">Compile PyTorch Object Detection Models</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_object_detection_pytorch.py</span></code>)</p></li>
+<li><p><strong>02:37.568</strong>: <a class="reference internal" href="deploy_ssd_gluoncv.html#sphx-glr-how-to-deploy-models-deploy-ssd-gluoncv-py"><span class="std std-ref">Deploy Single Shot Multibox Detector(SSD) model</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_ssd_gluoncv.py</span></code>)</p></li>
+<li><p><strong>02:14.102</strong>: <a class="reference internal" href="deploy_quantized.html#sphx-glr-how-to-deploy-models-deploy-quantized-py"><span class="std std-ref">Deploy a Quantized Model on Cuda</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_quantized.py</span></code>)</p></li>
+<li><p><strong>01:56.617</strong>: <a class="reference internal" href="deploy_prequantized_tflite.html#sphx-glr-how-to-deploy-models-deploy-prequantized-tflite-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM - Part 3 (TFLite)</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized_tflite.py</span></code>)</p></li>
+<li><p><strong>01:04.650</strong>: <a class="reference internal" href="deploy_prequantized.html#sphx-glr-how-to-deploy-models-deploy-prequantized-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized.py</span></code>)</p></li>
+<li><p><strong>00:27.652</strong>: <a class="reference internal" href="deploy_model_on_android.html#sphx-glr-how-to-deploy-models-deploy-model-on-android-py"><span class="std std-ref">Deploy the Pretrained Model on Android</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_android.py</span></code>)</p></li>
+<li><p><strong>00:21.858</strong>: <a class="reference internal" href="deploy_model_on_rasp.html#sphx-glr-how-to-deploy-models-deploy-model-on-rasp-py"><span class="std std-ref">Deploy the Pretrained Model on Raspberry Pi</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_rasp.py</span></code>)</p></li>
+<li><p><strong>00:00.206</strong>: <a class="reference internal" href="deploy_sparse.html#sphx-glr-how-to-deploy-models-deploy-sparse-py"><span class="std std-ref">Deploy a Hugging Face Pruned Model on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_sparse.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/extend_tvm/bring_your_own_datatypes.html b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
index 15c76fc2d..8be17f0f9 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -590,7 +590,7 @@ In this alpha state of the Bring Your Own Datatypes framework, we have not imple
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip448c3c6b-e2a1-468b-abfd-6265e0f34ccf from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip7a20ba62-ec3d-4d09-90ea-6fe5c24e8cab from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
</pre></div>
</div>
<p>It’s easy to execute MobileNet with native TVM:</p>
diff --git a/docs/how_to/extend_tvm/sg_execution_times.html b/docs/how_to/extend_tvm/sg_execution_times.html
index d2fd01bd0..51818ce48 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -300,12 +300,12 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-extend-tvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:37.164</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:37.575</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:33.792</strong>: <a class="reference internal" href="bring_your_own_datatypes.html#sphx-glr-how-to-extend-tvm-bring-your-own-datatypes-py"><span class="std std-ref">Bring Your Own Datatypes to TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">bring_your_own_datatypes.py</span></code>)</p></li>
-<li><p><strong>00:02.180</strong>: <a class="reference internal" href="use_pass_instrument.html#sphx-glr-how-to-extend-tvm-use-pass-instrument-py"><span class="std std-ref">How to Use TVM Pass Instrument</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_instrument.py</span></code>)</p></li>
-<li><p><strong>00:01.009</strong>: <a class="reference internal" href="use_pass_infra.html#sphx-glr-how-to-extend-tvm-use-pass-infra-py"><span class="std std-ref">How to Use TVM Pass Infra</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_infra.py</span></code>)</p></li>
-<li><p><strong>00:00.183</strong>: <a class="reference internal" href="low_level_custom_pass.html#sphx-glr-how-to-extend-tvm-low-level-custom-pass-py"><span class="std std-ref">Writing a Customized Pass</span></a> (<code class="docutils literal notranslate"><span class="pre">low_level_custom_pass.py</span></code>)</p></li>
+<li><p><strong>00:34.140</strong>: <a class="reference internal" href="bring_your_own_datatypes.html#sphx-glr-how-to-extend-tvm-bring-your-own-datatypes-py"><span class="std std-ref">Bring Your Own Datatypes to TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">bring_your_own_datatypes.py</span></code>)</p></li>
+<li><p><strong>00:02.204</strong>: <a class="reference internal" href="use_pass_instrument.html#sphx-glr-how-to-extend-tvm-use-pass-instrument-py"><span class="std std-ref">How to Use TVM Pass Instrument</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_instrument.py</span></code>)</p></li>
+<li><p><strong>00:01.023</strong>: <a class="reference internal" href="use_pass_infra.html#sphx-glr-how-to-extend-tvm-use-pass-infra-py"><span class="std std-ref">How to Use TVM Pass Infra</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_infra.py</span></code>)</p></li>
+<li><p><strong>00:00.207</strong>: <a class="reference internal" href="low_level_custom_pass.html#sphx-glr-how-to-extend-tvm-low-level-custom-pass-py"><span class="std std-ref">Writing a Customized Pass</span></a> (<code class="docutils literal notranslate"><span class="pre">low_level_custom_pass.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/extend_tvm/use_pass_instrument.html b/docs/how_to/extend_tvm/use_pass_instrument.html
index debba59c7..c2b6bdb06 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -486,10 +486,10 @@ profile the execution time of each passes.</p>
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 6103us [6103us] (45.64%; 45.64%)
-FoldScaleAxis: 7269us [6us] (54.36%; 54.36%)
- FoldConstant: 7263us [1477us] (54.32%; 99.92%)
- InferType: 5786us [5786us] (43.27%; 79.66%)
+InferType: 6028us [6028us] (45.49%; 45.49%)
+FoldScaleAxis: 7222us [6us] (54.51%; 54.51%)
+ FoldConstant: 7216us [1473us] (54.46%; 99.92%)
+ InferType: 5744us [5744us] (43.35%; 79.59%)
</pre></div>
</div>
</div>
@@ -512,10 +512,10 @@ Refer to following sections and <a class="reference internal" href="../../refere
</div>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 5825us [5825us] (44.52%; 44.52%)
-FoldScaleAxis: 7257us [5us] (55.48%; 55.48%)
- FoldConstant: 7253us [1517us] (55.44%; 99.94%)
- InferType: 5736us [5736us] (43.85%; 79.09%)
+InferType: 5810us [5810us] (44.63%; 44.63%)
+FoldScaleAxis: 7208us [5us] (55.37%; 55.37%)
+ FoldConstant: 7203us [1510us] (55.33%; 99.94%)
+ InferType: 5694us [5694us] (43.74%; 79.04%)
</pre></div>
</div>
<p>Register empty list to clear existing instruments.</p>
diff --git a/docs/how_to/optimize_operators/opt_conv_cuda.html b/docs/how_to/optimize_operators/opt_conv_cuda.html
index ce00efdc1..6a2a71e3d 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -534,7 +534,7 @@ latency of convolution.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.171636 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 37.823921 ms
</pre></div>
</div>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-optimize-operators-opt-conv-cuda-py">
diff --git a/docs/how_to/optimize_operators/opt_conv_tensorcore.html b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
index 2998223b4..f3cf080fa 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -878,7 +878,7 @@ be able to run on our build server</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 6.863368 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 9.411514 ms
</pre></div>
</div>
</div>
diff --git a/docs/how_to/optimize_operators/opt_gemm.html b/docs/how_to/optimize_operators/opt_gemm.html
index fc064e87e..3f14e061c 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -431,8 +431,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018139
-Baseline: 3.389786
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018413
+Baseline: 3.375813
</pre></div>
</div>
<p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -494,7 +494,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.291762
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.293211
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -563,7 +563,7 @@ vastly.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.331351
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.330585
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -626,7 +626,7 @@ the access pattern for A matrix is more cache friendly.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.113804
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.113394
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -711,7 +711,7 @@ flattening.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109979
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109488
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -799,7 +799,7 @@ write to C when all the block results are ready.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.111204
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.110785
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -891,7 +891,7 @@ write to C when all the block results are ready.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.144465
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.143222
</pre></div>
</div>
<p>Here is the generated IR after parallelization.</p>
diff --git a/docs/how_to/optimize_operators/sg_execution_times.html b/docs/how_to/optimize_operators/sg_execution_times.html
index fd4d8178b..0f7ed7434 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -300,11 +300,11 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:34.368</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:34.320</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:32.023</strong>: <a class="reference internal" href="opt_gemm.html#sphx-glr-how-to-optimize-operators-opt-gemm-py"><span class="std std-ref">How to optimize GEMM on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_gemm.py</span></code>)</p></li>
-<li><p><strong>00:01.266</strong>: <a class="reference internal" href="opt_conv_tensorcore.html#sphx-glr-how-to-optimize-operators-opt-conv-tensorcore-py"><span class="std std-ref">How to optimize convolution using TensorCores</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_tensorcore.py</span></code>)</p></li>
-<li><p><strong>00:01.079</strong>: <a class="reference internal" href="opt_conv_cuda.html#sphx-glr-how-to-optimize-operators-opt-conv-cuda-py"><span class="std std-ref">How to optimize convolution on GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_cuda.py</span></code>)</p></li>
+<li><p><strong>00:31.967</strong>: <a class="reference internal" href="opt_gemm.html#sphx-glr-how-to-optimize-operators-opt-gemm-py"><span class="std std-ref">How to optimize GEMM on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_gemm.py</span></code>)</p></li>
+<li><p><strong>00:01.303</strong>: <a class="reference internal" href="opt_conv_tensorcore.html#sphx-glr-how-to-optimize-operators-opt-conv-tensorcore-py"><span class="std std-ref">How to optimize convolution using TensorCores</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_tensorcore.py</span></code>)</p></li>
+<li><p><strong>00:01.051</strong>: <a class="reference internal" href="opt_conv_cuda.html#sphx-glr-how-to-optimize-operators-opt-conv-cuda-py"><span class="std std-ref">How to optimize convolution on GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_cuda.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
index 017206611..315a3e082 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -300,14 +300,14 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-tune-with-autoscheduler-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>03:28.794</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>03:29.523</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<ul class="simple">
-<li><p><strong>01:17.695</strong>: <a class="reference internal" href="tune_network_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-x86-py"><span class="std std-ref">Auto-scheduling a Neural Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_x86.py</span></code>)</p></li>
-<li><p><strong>01:05.687</strong>: <a class="reference internal" href="tune_conv2d_layer_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py"><span class="std std-ref">Auto-scheduling a Convolution Layer for GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_layer_cuda.py</span></code>)</p></li>
-<li><p><strong>00:39.971</strong>: <a class="reference internal" href="tune_network_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-cuda-py"><span class="std std-ref">Auto-scheduling a Neural Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_cuda.py</span></code>)</p></li>
-<li><p><strong>00:08.592</strong>: <a class="reference internal" href="tune_network_mali.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-mali-py"><span class="std std-ref">Auto-scheduling a Neural Network for mali GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_mali.py</span></code>)</p></li>
-<li><p><strong>00:08.541</strong>: <a class="reference internal" href="tune_sparse_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-sparse-x86-py"><span class="std std-ref">Auto-scheduling Sparse Matrix Multiplication on CPU with Custom Sketch Rule</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_sparse_x86.py</span></code>)</p></li>
-<li><p><strong>00:08.308</strong>: <a class="reference internal" href="tune_network_arm.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-arm-py"><span class="std std-ref">Auto-scheduling a Neural Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_arm.py</span></code>)</p></li>
+<li><p><strong>01:19.638</strong>: <a class="reference internal" href="tune_network_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-x86-py"><span class="std std-ref">Auto-scheduling a Neural Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_x86.py</span></code>)</p></li>
+<li><p><strong>01:02.355</strong>: <a class="reference internal" href="tune_conv2d_layer_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py"><span class="std std-ref">Auto-scheduling a Convolution Layer for GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_layer_cuda.py</span></code>)</p></li>
+<li><p><strong>00:40.302</strong>: <a class="reference internal" href="tune_network_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-cuda-py"><span class="std std-ref">Auto-scheduling a Neural Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_cuda.py</span></code>)</p></li>
+<li><p><strong>00:10.386</strong>: <a class="reference internal" href="tune_sparse_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-sparse-x86-py"><span class="std std-ref">Auto-scheduling Sparse Matrix Multiplication on CPU with Custom Sketch Rule</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_sparse_x86.py</span></code>)</p></li>
+<li><p><strong>00:08.549</strong>: <a class="reference internal" href="tune_network_mali.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-mali-py"><span class="std std-ref">Auto-scheduling a Neural Network for mali GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_mali.py</span></code>)</p></li>
+<li><p><strong>00:08.292</strong>: <a class="reference internal" href="tune_network_arm.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-arm-py"><span class="std std-ref">Auto-scheduling a Neural Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_arm.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html b/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
index f92a5ff7e..d24b7684f 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
@@ -470,12 +470,12 @@ cooperative fetching, unrolling and operator fusion.</p>
compute: Buffer(compute_2: Pointer(float32), float32, [25088], [])}
buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute}
preflattened_buffer_map = {data_1: data_3: Buffer(data_2, float32, [1, 512, 7, 7], []), kernel_1: kernel_3: Buffer(kernel_2, float32, [512, 512, 3, 3], []), bias_1: bias_3: Buffer(bias_2, float32, [1, 512, 1, 1], []), compute_1: compute_3: Buffer(compute_2, float32, [1, 512, 7, 7], [])} {
- attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 64;
- allocate(conv2d_nchw: Pointer(local float32), float32, [8]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [324]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [288]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [8], [], scope="local", align=32)[0] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 28;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [14], [], scope="local", align=32)[0] = 0f32
conv2d_nchw_1[1] = 0f32
conv2d_nchw_1[2] = 0f32
conv2d_nchw_1[3] = 0f32
@@ -483,88 +483,470 @@ cooperative fetching, unrolling and operator fusion.</p>
conv2d_nchw_1[5] = 0f32
conv2d_nchw_1[6] = 0f32
conv2d_nchw_1[7] = 0f32
- for (rc.outer.outer: int32, 0, 128) {
- let cse_var_1: int32 = (rc.outer.outer*196)
- {
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [324], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((9 <= threadIdx.x_1) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[(((cse_var_1 + (floordiv(threadIdx.x_1, 9)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 49)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 49), 81)) && (floormod((threadIdx.x_1 + 49), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 49), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 49), 81), 9)*7)) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 98)] = @tir.if_then_else(((1 <= floormod((threadIdx.x_1 + 8), 9)) && (floormod((threadIdx.x_1 + 8), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 98), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 98), 81), 9)*7)) + floormod((threadIdx.x_1 + 8), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 147)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 147), 81)) && (floormod((threadIdx.x_1 + 66), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 147), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 147), 81), 9)*7)) + floormod((threadIdx.x_1 + 3), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 196)] = @tir.if_then_else(((((9 <= floormod((threadIdx.x_1 + 196), 81)) && (floormod((threadIdx.x_1 + 34), 81) < 72)) && (1 <= floormod((threadIdx.x_1 + 7), 9))) && (floormod((threadIdx.x_1 + 7), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 196), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 196), 81), 9)*7)) + floormod((threadIdx.x_1 + 7), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- pad_temp.shared_1[(threadIdx.x_1 + 245)] = @tir.if_then_else((((9 <= floormod((threadIdx.x_1 + 245), 81)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 245), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 245), 81), 9)*7)) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
- if @tir.likely((threadIdx.x_1 < 30), dtype=bool) {
- pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else((((floormod((threadIdx.x_1 + 51), 81) < 72) && (1 <= floormod((threadIdx.x_1 + 6), 9))) && (floormod((threadIdx.x_1 + 6), 9) < 8)), data[((((cse_var_1 + (floordiv((threadIdx.x_1 + 294), 81)*49)) + (floordiv(floormod((threadIdx.x_1 + 294), 81), 9)*7)) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
- }
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1: Buffer(kernel.shared, float32, [288], [], scope="shared")[(threadIdx.x_2*8)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod((threadIdx.x_2*8), 36), 3)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 1)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 1), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 2)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 2), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 3)] = kernel[(((((blockIdx.x*36864) + (floordiv((threadIdx.x_2*2), 9)*4608)) + (rc.outer.outer*36)) + (floormod((floordiv((threadIdx.x_2*8), 3) + 1), 12)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 4)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 4), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 5)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 5), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 6)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floormod((floordiv((threadIdx.x_2*8), 3) + 2), 12)*3)) + floormod((threadIdx.x_2*2), 3))]
- }
- if @tir.likely((threadIdx.x_2 < 36), dtype=bool) {
- kernel.shared_1[((threadIdx.x_2*8) + 7)] = kernel[(((((blockIdx.x*36864) + (floordiv(((threadIdx.x_2*2) + 1), 9)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(((threadIdx.x_2*8) + 7), 36), 3)*3)) + floormod(((threadIdx.x_2*2) + 1), 3))]
- }
- }
- for (rx.outer.inner: int32, 0, 3) {
- for (ff.outer.inner: int32, 0, 4) {
- let cse_var_4: int32 = (ff.outer.inner*2)
- let cse_var_3: int32 = ((ff.outer.inner*72) + rx.outer.inner)
- let cse_var_2: int32 = (cse_var_4 + 1)
- {
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[cse_var_3]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[(((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_3 + 36)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[(cse_var_3 + 3)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 9)]*kernel.shared_1[(cse_var_3 + 39)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[(cse_var_3 + 6)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 18)]*kernel.shared_1[(cse_var_3 + 42)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[(cse_var_3 + 9)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 81)]*kernel.shared_1[(cse_var_3 + 45)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[(cse_var_3 + 12)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 90)]*kernel.shared_1[(cse_var_3 + 48)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[(cse_var_3 + 15)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 99)]*kernel.shared_1[(cse_var_3 + 51)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[(cse_var_3 + 18)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 162)]*kernel.shared_1[(cse_var_3 + 54)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[(cse_var_3 + 21)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 171)]*kernel.shared_1[(cse_var_3 + 57)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[(cse_var_3 + 24)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 180)]*kernel.shared_1[(cse_var_3 + 60)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[(cse_var_3 + 27)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 243)]*kernel.shared_1[(cse_var_3 + 63)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[(cse_var_3 + 30)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 252)]*kernel.shared_1[(cse_var_3 + 66)]))
- conv2d_nchw_1[cse_var_4] = (conv2d_nchw_1[cse_var_4] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[(cse_var_3 + 33)]))
- conv2d_nchw_1[cse_var_2] = (conv2d_nchw_1[cse_var_2] + (pad_temp.shared_1[((((floordiv(threadIdx.x, 7)*9) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 261)]*kernel.shared_1[(cse_var_3 + 69)]))
+ conv2d_nchw_1[8] = 0f32
+ conv2d_nchw_1[9] = 0f32
+ conv2d_nchw_1[10] = 0f32
+ conv2d_nchw_1[11] = 0f32
+ conv2d_nchw_1[12] = 0f32
+ conv2d_nchw_1[13] = 0f32
+ for (rc.outer.outer: int32, 0, 64) {
+ for (ry.outer.outer: int32, 0, 3) {
+ let cse_var_2: int32 = (rc.outer.outer*72)
+ let cse_var_1: int32 = (ry.outer.outer*3)
+ {
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope="shared")[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1*4), 9))) && (floormod((threadIdx.x_1*4), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv((threadIdx.x_1*4), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 1), 9))) && (floormod(((threadIdx.x_1*4) + 1), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 1), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 1), 9)) - 8)], 0 [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 2), 9))) && (floormod(((threadIdx.x_1*4) + 2), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 2), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 2), 9)) - 8)], 0 [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 3), 9))) && (floormod(((threadIdx.x_1*4) + 3), 9) < 8)), data[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 3), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 3), 9)) - 8)], 0 [...]
}
}
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope="shared")[threadIdx.x_2] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 64)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 8), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 128)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 16), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 32), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 192)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 36864)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 256)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 32), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 64), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 320)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 40), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 80), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 384)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 73728)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 448)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 56), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 112), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 512)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 64), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 128), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 576)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 110592)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 640)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 80), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 160), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 704)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 88), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 176), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 768)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 147456)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 832)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 104), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 208), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 896)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 112), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 224), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 960)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 184320)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 128), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 256), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 136), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 272), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 221184)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 152), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 304), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 160), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 320), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 258048)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 176), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 352), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 184), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 368), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 294912)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 200), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 400), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 208), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 416), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 331776)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 224), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 448), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 232), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 464), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 368640)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 248), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 496), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 256), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 512), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 405504)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 272), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 544), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 280), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 560), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 442368)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 296), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 592), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 304), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 608), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 479232)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 320), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 640), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 328), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 656), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 516096)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 344), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 688), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 352), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 704), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(floordiv(threadIdx.x_2, 8), 3)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 552960)]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 368), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 736), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+ kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((floordiv(threadIdx.x_2, 8) + 376), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 752), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 3)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 24)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 27)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 4)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 25)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 28)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 5)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 26)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 29)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 6)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 9)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 30)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 33)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 7)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 10)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 31)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 34)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 8)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 11)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 32)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 35)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 12)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 15)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 36)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 39)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 13)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 16)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 37)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 40)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 14)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 17)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 38)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 41)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 18)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 21)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 42)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 45)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 19)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 22)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 43)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 46)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 20)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 23)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 44)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 47)]))
}
}
}
- for (i1.inner: int32, 0, 8) {
- compute[(((blockIdx.x*392) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias[((blockIdx.x*8) + i1.inner)]), 0f32)
+ for (i1.inner: int32, 0, 2) {
+ for (i3.inner: int32, 0, 7) {
+ compute[(((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*98)) + (i1.inner*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[((i1.inner*7) + i3.inner)] + bias[(((floordiv(blockIdx.x, 7)*128) + (threadIdx.x*2)) + i1.inner)]), 0f32)
+ }
}
}
}
@@ -602,7 +984,7 @@ cooperative fetching, unrolling and operator fusion.</p>
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.311 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.365 ms
</pre></div>
</div>
</div>
@@ -632,21 +1014,21 @@ conv2d_nchw_nn_o_i, conv2d_nchw_nn_i = s[conv2d_nchw].split(conv2d_nchw_nn, fact
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
-conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
-conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=4)
-conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=1)
+conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
+conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=64)
conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
-conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
-conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
-conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=7)
+conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=7)
+conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
-conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=4)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
-conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
+conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=2)
+conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
+conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
@@ -654,14 +1036,14 @@ s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nc
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=8)
-compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=1)
+compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
+compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=64)
compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
-compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
-compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
-compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
+compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
+compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
@@ -679,16 +1061,16 @@ s[compute].bind(compute_i0_o_o_i_i1_o_o_i_fused_i2_o_o_i_fused_i3_o_o_i_fused, t
compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused = s[compute].fuse(compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i)
s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread_axis("threadIdx.x"))
kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[kernel_shared].fuse(kernel_shared_ax0, kernel_shared_ax1, kernel_shared_ax2, kernel_shared_ax3)
-kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=8)
+kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
-kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
+kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=49)
+pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
-s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 64)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -706,10 +1088,10 @@ CUDA source code:
#define int64_t long long
#define uint64_t unsigned long long
#endif
-extern "C" __global__ void __launch_bounds__(49) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[8];
- __shared__ float pad_temp_shared[324];
- __shared__ float kernel_shared[288];
+extern "C" __global__ void __launch_bounds__(64) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[14];
+ __shared__ float pad_temp_shared[72];
+ __shared__ float kernel_shared[3072];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
conv2d_nchw[2] = 0.000000e+00f;
@@ -718,73 +1100,418 @@ extern "C" __global__ void __launch_bounds__(49) default_function_kern
conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[7] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
- __syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = ((((9 <= ((int)threadIdx.x)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((rc_outer_outer * 196) + ((((int)threadIdx.x) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((9 <= ((((int)threadIdx.x) + 49) % 81)) && (((((int)threadIdx.x) + 49) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 49) / 81) * 49)) + ((((((int)threadIdx.x) + 49) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 98)] = (((1 <= ((((int)threadIdx.x) + 8) % 9)) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 98) / 81) * 49)) + ((((((int)threadIdx.x) + 17) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((9 <= ((((int)threadIdx.x) + 66) % 81)) && (((((int)threadIdx.x) + 66) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 147) / 81) * 49)) + ((((((int)threadIdx.x) + 66) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((9 <= ((((int)threadIdx.x) + 34) % 81)) && (((((int)threadIdx.x) + 34) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 196) / 81) * 49)) + ((((((int)threadIdx.x) + 34) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 245)] = ((((9 <= ((((int)threadIdx.x) + 2) % 81)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 245) / 81) * 49)) + ((((((int)threadIdx.x) + 2) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
- if (((int)threadIdx.x) < 30) {
- pad_temp_shared[(((int)threadIdx.x) + 294)] = ((((((int)threadIdx.x) < 21) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) + 294) / 81) * 49)) + ((((((int)threadIdx.x) + 51) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[(((int)threadIdx.x) * 8)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + ((((((int)threadIdx.x) * 8) % 36) / 3) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 1)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 1) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 2)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 2) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 3)] = kernel[(((((((int)blockIdx.x) * 36864) + (((((int)threadIdx.x) * 2) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) / 3) + 1) % 12) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 4)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 4) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 5)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 5) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 6)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) / 3) + 2) % 12) * 3)) + ((((int)threadIdx.x) * 2) % 3))];
- }
- if (((int)threadIdx.x) < 36) {
- kernel_shared[((((int)threadIdx.x) * 8) + 7)] = kernel[(((((((int)blockIdx.x) * 36864) + ((((((int)threadIdx.x) * 2) + 1) / 9) * 4608)) + (rc_outer_outer * 36)) + (((((((int)threadIdx.x) * 8) + 7) % 36) / 3) * 3)) + (((((int)threadIdx.x) * 2) + 1) % 3))];
- }
- __syncthreads();
- for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
- for (int ff_outer_inner = 0; ff_outer_inner < 4; ++ff_outer_inner) {
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[((ff_outer_inner * 72) + rx_outer_inner)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 36)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 3)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 9)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 39)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 6)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 18)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 42)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 9)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 81)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 45)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 12)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 90)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 48)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 15)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 99)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 51)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 18)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 162)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 54)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 21)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 171)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 57)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 24)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 180)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 60)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 27)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 243)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 63)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 30)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 252)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 66)]));
- conv2d_nchw[(ff_outer_inner * 2)] = (conv2d_nchw[(ff_outer_inner * 2)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 33)]));
- conv2d_nchw[((ff_outer_inner * 2) + 1)] = (conv2d_nchw[((ff_outer_inner * 2) + 1)] + (pad_temp_shared[(((((((int)threadIdx.x) / 7) * 9) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 261)] * kernel_shared[(((ff_outer_inner * 72) + rx_outer_inner) + 69)]));
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[13] = 0.000000e+00f;
+ for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
+ for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
+ __syncthreads();
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) * 4) % 9))) && (((((int)threadIdx.x) * 4) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 4) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 1) % 9))) && ((((((int)threadIdx.x) * 4) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.000000e+00f);
}
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 2) % 9))) && ((((((int)threadIdx.x) * 4) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 18) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 3) % 9))) && ((((((int)threadIdx.x) * 4) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.000000e+00f);
+ }
+ kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 36864)];
+ kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 73728)];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 110592)];
+ kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 147456)];
+ kernel_shared[(((int)threadIdx.x) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 184320)];
+ kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 221184)];
+ kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 258048)];
+ kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 294912)];
+ kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 331776)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 368640)];
+ kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 405504)];
+ kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 442368)];
+ kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 479232)];
+ kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 516096)];
+ kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 552960)];
+ kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ __syncthreads();
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
}
}
- for (int i1_inner = 0; i1_inner < 8; ++i1_inner) {
- compute[(((((int)blockIdx.x) * 392) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 8) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[((((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[((((((int)blockIdx.x) / 7) * 128) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+ }
}
}
</pre></div>
@@ -822,7 +1549,7 @@ In the example below we resume the status and do more 5 trials.</p>
Get devices for measurement successfully!
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 5.687 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 2.355 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/e3e540f3b477c0c52d8eb73e674e8ffd/tune_conv2d_layer_cuda.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">tune_conv2d_layer_cuda.py</span></code></a></p>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
index b74251522..1e2a2c7d1 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -878,7 +878,7 @@ so we can read the log file and load the best schedules.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 9.9082 9.9383 9.9670 9.8194 0.0639
+ 9.9023 9.8786 9.9577 9.8707 0.0392
</pre></div>
</div>
</div>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
index ab54ba403..66e67555e 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -897,7 +897,7 @@ so we can read the log file and load the best schedules.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 749.4805 749.3528 750.0807 749.0080 0.4471
+ 804.0846 803.9296 804.5431 803.7811 0.3298
</pre></div>
</div>
</div>
@@ -919,7 +919,7 @@ to learn how to use the RPC Tracker and RPC Server.
To use the RPC Tracker in auto-scheduler, replace the runner in <code class="code docutils literal notranslate"><span class="pre">TuningOptions</span></code>
with <a class="reference internal" href="../../reference/api/python/auto_scheduler.html#tvm.auto_scheduler.RPCRunner" title="tvm.auto_scheduler.RPCRunner"><code class="xref any py py-class docutils literal notranslate"><span class="pre">auto_scheduler.RPCRunner</span></code></a>.</p></li>
</ol>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 17.695 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 19.638 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autoscheduler-tune-network-x86-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/e416b94ca1090b0897c0f6e0df95b911/tune_network_x86.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">tune_network_x86.py</span></code></a></p>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
index 236abdb91..77cd30e4e 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -600,71 +600,69 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [65536], []),
compute: Buffer(compute_2: Pointer(float32), float32, [65536], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute}
- preflattened_buffer_map = {placeholder_5: placeholder_15: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], []), placeholder_6: placeholder_16: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_9: placeholder_17: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_18: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_19: Buffer(placeholder_12, int32, [4916], [])} {
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_9: placeholder_16: Buffer(placeholder_14, float32, [128, 512], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_5: placeholder_19: Buffer(placeholder_10, float32, [128, 256], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
for (i0.outer.i1.outer.fused: int32, 0, 32) "parallel" {
allocate(compute_4: Pointer(global float32), float32, [2048]), storage_scope = global {
- for (i.outer.inner: int32, 0, 2) {
- for (nb_j.inner: int32, 0, 2) {
- for (i.inner.init: int32, 0, 32) {
- let cse_var_1: int32 = (((i.outer.inner*1024) + (i.inner.init*32)) + (nb_j.inner*16))
- {
- compute_5: Buffer(compute_4, float32, [2048], [])[cse_var_1] = 0f32
- compute_5[(cse_var_1 + 1)] = 0f32
- compute_5[(cse_var_1 + 2)] = 0f32
- compute_5[(cse_var_1 + 3)] = 0f32
- compute_5[(cse_var_1 + 4)] = 0f32
- compute_5[(cse_var_1 + 5)] = 0f32
- compute_5[(cse_var_1 + 6)] = 0f32
- compute_5[(cse_var_1 + 7)] = 0f32
- compute_5[(cse_var_1 + 8)] = 0f32
- compute_5[(cse_var_1 + 9)] = 0f32
- compute_5[(cse_var_1 + 10)] = 0f32
- compute_5[(cse_var_1 + 11)] = 0f32
- compute_5[(cse_var_1 + 12)] = 0f32
- compute_5[(cse_var_1 + 13)] = 0f32
- compute_5[(cse_var_1 + 14)] = 0f32
- compute_5[(cse_var_1 + 15)] = 0f32
- }
+ for (nb_j.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 64) {
+ let cse_var_1: int32 = ((i.inner.init*32) + (nb_j.inner*16))
+ {
+ compute_5: Buffer(compute_4, float32, [2048], [])[cse_var_1] = 0f32
+ compute_5[(cse_var_1 + 1)] = 0f32
+ compute_5[(cse_var_1 + 2)] = 0f32
+ compute_5[(cse_var_1 + 3)] = 0f32
+ compute_5[(cse_var_1 + 4)] = 0f32
+ compute_5[(cse_var_1 + 5)] = 0f32
+ compute_5[(cse_var_1 + 6)] = 0f32
+ compute_5[(cse_var_1 + 7)] = 0f32
+ compute_5[(cse_var_1 + 8)] = 0f32
+ compute_5[(cse_var_1 + 9)] = 0f32
+ compute_5[(cse_var_1 + 10)] = 0f32
+ compute_5[(cse_var_1 + 11)] = 0f32
+ compute_5[(cse_var_1 + 12)] = 0f32
+ compute_5[(cse_var_1 + 13)] = 0f32
+ compute_5[(cse_var_1 + 14)] = 0f32
+ compute_5[(cse_var_1 + 15)] = 0f32
}
- for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
- for (i.inner: int32, 0, 32) {
- let cse_var_21: int32 = (elem_idx*16)
- let cse_var_20: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
- let cse_var_19: int32 = (((i.outer.inner*1024) + (i.inner*32)) + (nb_j.inner*16))
- let cse_var_18: int32 = (cse_var_19 + 1)
- let cse_var_17: int32 = (cse_var_19 + 11)
- let cse_var_16: int32 = (cse_var_19 + 12)
- let cse_var_15: int32 = (cse_var_19 + 13)
- let cse_var_14: int32 = (cse_var_19 + 14)
- let cse_var_13: int32 = (cse_var_19 + 15)
- let cse_var_12: int32 = (cse_var_19 + 2)
- let cse_var_11: int32 = (cse_var_19 + 3)
- let cse_var_10: int32 = (cse_var_19 + 4)
- let cse_var_9: int32 = (cse_var_19 + 5)
- let cse_var_8: int32 = (cse_var_19 + 6)
- let cse_var_7: int32 = (cse_var_19 + 7)
- let cse_var_6: int32 = (cse_var_19 + 8)
- let cse_var_5: int32 = (cse_var_19 + 9)
- let cse_var_4: int32 = (((floordiv(i0.outer.i1.outer.fused, 16)*16384) + (i.outer.inner*8192)) + (i.inner*256))
- let cse_var_3: int32 = (cse_var_19 + 10)
- {
- compute_5[cse_var_19] = (compute_5[cse_var_19] + (placeholder_1[((placeholder_3[cse_var_20]*16) + cse_var_21)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 1)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 2)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 3)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 4)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 5)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 6)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 7)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 8)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 9)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 10)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 11)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 12)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 13)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 14)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_20]*16) + cse_var_21) + 15)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_20] + elem_idx)])], 0f32)))
- }
+ }
+ for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
+ for (i.inner: int32, 0, 64) {
+ let cse_var_21: int32 = (elem_idx*16)
+ let cse_var_20: int32 = ((i.inner*32) + (nb_j.inner*16))
+ let cse_var_19: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
+ let cse_var_18: int32 = (cse_var_20 + 1)
+ let cse_var_17: int32 = (cse_var_20 + 11)
+ let cse_var_16: int32 = (cse_var_20 + 12)
+ let cse_var_15: int32 = (cse_var_20 + 13)
+ let cse_var_14: int32 = (cse_var_20 + 14)
+ let cse_var_13: int32 = (cse_var_20 + 15)
+ let cse_var_12: int32 = (cse_var_20 + 2)
+ let cse_var_11: int32 = (cse_var_20 + 3)
+ let cse_var_10: int32 = (cse_var_20 + 4)
+ let cse_var_9: int32 = (cse_var_20 + 5)
+ let cse_var_8: int32 = (cse_var_20 + 6)
+ let cse_var_7: int32 = (cse_var_20 + 7)
+ let cse_var_6: int32 = (cse_var_20 + 8)
+ let cse_var_5: int32 = (cse_var_20 + 9)
+ let cse_var_4: int32 = ((floordiv(i0.outer.i1.outer.fused, 16)*16384) + (i.inner*256))
+ let cse_var_3: int32 = (cse_var_20 + 10)
+ {
+ compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[((placeholder_3[cse_var_19]*16) + cse_var_21)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 1)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 2)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 3)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 4)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 5)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 6)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 7)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 8)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 9)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_3] = (compute_5[cse_var_3] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 10)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 11)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 12)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 13)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 14)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
+ compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_19]*16) + cse_var_21) + 15)]*max(placeholder[(cse_var_4 + placeholder_2[(placeholder_3[cse_var_19] + elem_idx)])], 0f32)))
}
}
}
@@ -710,7 +708,7 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
</pre></div>
</div>
<p class="sphx-glr-script-out">Out:</p>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.724 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.837 ms
</pre></div>
</div>
<div class="admonition note">
diff --git a/docs/how_to/tune_with_autotvm/sg_execution_times.html b/docs/how_to/tune_with_autotvm/sg_execution_times.html
index e65b5deaf..e61251836 100644
--- a/docs/how_to/tune_with_autotvm/sg_execution_times.html
+++ b/docs/how_to/tune_with_autotvm/sg_execution_times.html
@@ -300,13 +300,13 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-tune-with-autotvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:45.186</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:45.261</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:44.398</strong>: <a class="reference internal" href="tune_conv2d_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-conv2d-cuda-py"><span class="std std-ref">Tuning High Performance Convolution on NVIDIA GPUs</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_cuda.py</span></code>)</p></li>
-<li><p><strong>00:00.206</strong>: <a class="reference internal" href="tune_relay_x86.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-x86-py"><span class="std std-ref">Auto-tuning a Convolutional Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_x86.py</span></code>)</p></li>
-<li><p><strong>00:00.196</strong>: <a class="reference internal" href="tune_relay_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-cuda-py"><span class="std std-ref">Auto-tuning a Convolutional Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_cuda.py</span></code>)</p></li>
-<li><p><strong>00:00.194</strong>: <a class="reference internal" href="tune_relay_arm.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-arm-py"><span class="std std-ref">Auto-tuning a Convolutional Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_arm.py</span></code>)</p></li>
-<li><p><strong>00:00.191</strong>: <a class="reference internal" href="tune_relay_mobile_gpu.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-mobile-gpu-py"><span class="std std-ref">Auto-tuning a Convolutional Network for Mobile GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_mobile_gpu.py</span></code>)</p></li>
+<li><p><strong>00:44.454</strong>: <a class="reference internal" href="tune_conv2d_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-conv2d-cuda-py"><span class="std std-ref">Tuning High Performance Convolution on NVIDIA GPUs</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_cuda.py</span></code>)</p></li>
+<li><p><strong>00:00.219</strong>: <a class="reference internal" href="tune_relay_mobile_gpu.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-mobile-gpu-py"><span class="std std-ref">Auto-tuning a Convolutional Network for Mobile GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_mobile_gpu.py</span></code>)</p></li>
+<li><p><strong>00:00.204</strong>: <a class="reference internal" href="tune_relay_x86.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-x86-py"><span class="std std-ref">Auto-tuning a Convolutional Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_x86.py</span></code>)</p></li>
+<li><p><strong>00:00.192</strong>: <a class="reference internal" href="tune_relay_arm.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-arm-py"><span class="std std-ref">Auto-tuning a Convolutional Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_arm.py</span></code>)</p></li>
+<li><p><strong>00:00.191</strong>: <a class="reference internal" href="tune_relay_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-cuda-py"><span class="std std-ref">Auto-tuning a Convolutional Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_cuda.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
index 00a18501c..b64e181de 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -1142,8 +1142,8 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2885496
-No: 6 GFLOPS: 111.00/111.00 result: MeasureResult(costs=(0.002085639479166667,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.2219181060791016, timestamp=1654041265.559902) [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3754080
-No: 7 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 6 GFLOPS: 109.94/109.94 result: MeasureResult(costs=(0.0021056749375,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.230674982070923, timestamp=1654091865.1226287) [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3754080
+No: 7 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1266,7 +1266,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 16, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6225319
-No: 8 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 8 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1389,7 +1389,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,943546
-No: 9 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 9 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1512,7 +1512,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2868708
-No: 10 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 10 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
res = future.result()
File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
@@ -1530,7 +1530,7 @@ No: 10 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
TimeoutError
[('tile_f', [-1, 32, 2, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4691833
-No: 11 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 11 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1653,7 +1653,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 2, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1042124
-No: 12 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 12 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1776,7 +1776,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10013405
-No: 13 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 13 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -1899,7 +1899,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 8, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6732082
-No: 14 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 14 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2022,7 +2022,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7536735
-No: 15 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 15 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2145,7 +2145,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,482121
-No: 16 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 16 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2268,7 +2268,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2824525
-No: 17 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 17 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2391,7 +2391,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4559286
-No: 18 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 18 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 571, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 523, in _build_func_common
@@ -2514,7 +2514,7 @@ Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 854, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9677544
-No: 19 GFLOPS: 0.00/111.00 result: Traceback (most recent call last):
+No: 19 GFLOPS: 0.00/109.94 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 721, in __call__
yield remote, remote.load_module(os.path.split(build_result.filename)[1])
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 685, in run_through_rpc
@@ -2602,7 +2602,7 @@ tvm._ffi.base.TVMError: Traceback (most recent call last):
15: _PyEval_EvalFrameDefault
14: 0x0000000000537c30
13: _PyObject_FastCallKeywords
- 12: 0x00007f341f4e3fa2
+ 12: 0x00007f1ae6d0bfa2
11: _ctypes_callproc
10: ffi_call
9: ffi_call_unix64
@@ -2667,7 +2667,7 @@ Traceback (most recent call last):
21: _PyFunction_FastCallKeywords
20: _PyEval_EvalFrameDefault
19: _PyFunction_FastCall [('tile_f', [-1, 8, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6390073
-No: 20 GFLOPS: 143.96/143.96 result: MeasureResult(costs=(0.00160804004,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.390920877456665, timestamp=1654041283.8257668) [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
+No: 20 GFLOPS: 143.88/143.88 result: MeasureResult(costs=(0.0016089895300000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3828744888305664, timestamp=1654091884.3543549) [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
</pre></div>
</div>
<p>Finally we can inspect the best config from log file, check correctness,
@@ -2706,7 +2706,7 @@ and measure running time.</p>
<p class="sphx-glr-script-out">Out:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Best config:
[('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9881539
-Time cost of this operator: 0.001975
+Time cost of this operator: 0.001973
</pre></div>
</div>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autotvm-tune-conv2d-cuda-py">
diff --git a/docs/how_to/work_with_microtvm/micro_autotune.html b/docs/how_to/work_with_microtvm/micro_autotune.html
index fc990f6b3..54d3ffa33 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -555,10 +555,10 @@ the tuned operator.</p>
########## Build without Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs
--------- --- -------- ------- ----- ------ -------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 311.5 98.763 (1, 2, 10, 10, 3) 2 1
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.951 (1, 6, 10, 10) 1 1
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 0.286 (1, 1, 10, 10, 3) 1 1
-Total_time - 315.401 - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.1 98.758 (1, 2, 10, 10, 3) 2 1
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.955 (1, 6, 10, 10) 1 1
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 0.287 (1, 1, 10, 10, 3) 1 1
+Total_time - 314.001 - - - -
</pre></div>
</div>
</div>
@@ -610,10 +610,10 @@ Total_time -
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>########## Build with Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs
--------- --- -------- ------- ----- ------ -------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 198.7 98.744 (1, 6, 10, 10, 1) 2 1
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.728 0.859 (1, 6, 10, 10) 1 1
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.8 0.398 (1, 3, 10, 10, 1) 1 1
-Total_time - 201.228 - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 78.45 96.779 (1, 6, 10, 10, 1) 2 1
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.71 2.11 (1, 6, 10, 10) 1 1
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.901 1.111 (1, 1, 10, 10, 3) 1 1
+Total_time - 81.061 - - - -
</pre></div>
</div>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-work-with-microtvm-micro-autotune-py">
diff --git a/docs/how_to/work_with_microtvm/sg_execution_times.html b/docs/how_to/work_with_microtvm/sg_execution_times.html
index 4676aeef9..5c28ffb0c 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -300,12 +300,12 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-microtvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:45.415</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>00:46.313</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:41.463</strong>: <a class="reference internal" href="micro_autotune.html#sphx-glr-how-to-work-with-microtvm-micro-autotune-py"><span class="std std-ref">Autotuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></li>
-<li><p><strong>00:03.403</strong>: <a class="reference internal" href="micro_tflite.html#sphx-glr-how-to-work-with-microtvm-micro-tflite-py"><span class="std std-ref">microTVM with TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></li>
-<li><p><strong>00:00.189</strong>: <a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></li>
-<li><p><strong>00:00.183</strong>: <a class="reference internal" href="micro_tvmc.html#sphx-glr-how-to-work-with-microtvm-micro-tvmc-py"><span class="std std-ref">Executing a Tiny Model with TVMC Micro</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tvmc.py</span></code>)</p></li>
+<li><p><strong>00:42.355</strong>: <a class="reference internal" href="micro_autotune.html#sphx-glr-how-to-work-with-microtvm-micro-autotune-py"><span class="std std-ref">Autotuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></li>
+<li><p><strong>00:03.406</strong>: <a class="reference internal" href="micro_tflite.html#sphx-glr-how-to-work-with-microtvm-micro-tflite-py"><span class="std std-ref">microTVM with TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></li>
+<li><p><strong>00:00.190</strong>: <a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></li>
+<li><p><strong>00:00.185</strong>: <a class="reference internal" href="micro_tvmc.html#sphx-glr-how-to-work-with-microtvm-micro-tvmc-py"><span class="std std-ref">Executing a Tiny Model with TVMC Micro</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tvmc.py</span></code>)</p></li>
<li><p><strong>00:00.177</strong>: <a class="reference internal" href="micro_reference_vm.html#sphx-glr-how-to-work-with-microtvm-micro-reference-vm-py"><span class="std std-ref">microTVM Reference Virtual Machines</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_reference_vm.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/work_with_relay/sg_execution_times.html b/docs/how_to/work_with_relay/sg_execution_times.html
index 704589130..d7a598e8c 100644
--- a/docs/how_to/work_with_relay/sg_execution_times.html
+++ b/docs/how_to/work_with_relay/sg_execution_times.html
@@ -300,11 +300,11 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-relay-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:05.015</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:05.058</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:03.423</strong>: <a class="reference internal" href="using_external_lib.html#sphx-glr-how-to-work-with-relay-using-external-lib-py"><span class="std std-ref">Using External Libraries in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_external_lib.py</span></code>)</p></li>
-<li><p><strong>00:01.381</strong>: <a class="reference internal" href="build_gcn.html#sphx-glr-how-to-work-with-relay-build-gcn-py"><span class="std std-ref">Building a Graph Convolutional Network</span></a> (<code class="docutils literal notranslate"><span class="pre">build_gcn.py</span></code>)</p></li>
-<li><p><strong>00:00.210</strong>: <a class="reference internal" href="using_relay_viz.html#sphx-glr-how-to-work-with-relay-using-relay-viz-py"><span class="std std-ref">Use Relay Visualizer to Visualize Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_relay_viz.py</span></code>)</p></li>
+<li><p><strong>00:03.438</strong>: <a class="reference internal" href="using_external_lib.html#sphx-glr-how-to-work-with-relay-using-external-lib-py"><span class="std std-ref">Using External Libraries in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_external_lib.py</span></code>)</p></li>
+<li><p><strong>00:01.408</strong>: <a class="reference internal" href="build_gcn.html#sphx-glr-how-to-work-with-relay-build-gcn-py"><span class="std std-ref">Building a Graph Convolutional Network</span></a> (<code class="docutils literal notranslate"><span class="pre">build_gcn.py</span></code>)</p></li>
+<li><p><strong>00:00.212</strong>: <a class="reference internal" href="using_relay_viz.html#sphx-glr-how-to-work-with-relay-using-relay-viz-py"><span class="std std-ref">Use Relay Visualizer to Visualize Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_relay_viz.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/work_with_schedules/sg_execution_times.html b/docs/how_to/work_with_schedules/sg_execution_times.html
index bab2c2d5d..97aa238ec 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -300,16 +300,16 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-work-with-schedules-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:04.940</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:04.922</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:01.864</strong>: <a class="reference internal" href="intrin_math.html#sphx-glr-how-to-work-with-schedules-intrin-math-py"><span class="std std-ref">Intrinsics and Math Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">intrin_math.py</span></code>)</p></li>
-<li><p><strong>00:00.773</strong>: <a class="reference internal" href="tensorize.html#sphx-glr-how-to-work-with-schedules-tensorize-py"><span class="std std-ref">Use Tensorize to Leverage Hardware Intrinsics</span></a> (<code class="docutils literal notranslate"><span class="pre">tensorize.py</span></code>)</p></li>
-<li><p><strong>00:00.669</strong>: <a class="reference internal" href="reduction.html#sphx-glr-how-to-work-with-schedules-reduction-py"><span class="std std-ref">Reduction</span></a> (<code class="docutils literal notranslate"><span class="pre">reduction.py</span></code>)</p></li>
-<li><p><strong>00:00.664</strong>: <a class="reference internal" href="scan.html#sphx-glr-how-to-work-with-schedules-scan-py"><span class="std std-ref">Scan and Recurrent Kernel</span></a> (<code class="docutils literal notranslate"><span class="pre">scan.py</span></code>)</p></li>
-<li><p><strong>00:00.302</strong>: <a class="reference internal" href="extern_op.html#sphx-glr-how-to-work-with-schedules-extern-op-py"><span class="std std-ref">External Tensor Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">extern_op.py</span></code>)</p></li>
-<li><p><strong>00:00.233</strong>: <a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></li>
-<li><p><strong>00:00.224</strong>: <a class="reference internal" href="tedd.html#sphx-glr-how-to-work-with-schedules-tedd-py"><span class="std std-ref">Use Tensor Expression Debug Display (TEDD) for Visualization</span></a> (<code class="docutils literal notranslate"><span class="pre">tedd.py</span></code>)</p></li>
-<li><p><strong>00:00.210</strong>: <a class="reference internal" href="tuple_inputs.html#sphx-glr-how-to-work-with-schedules-tuple-inputs-py"><span class="std std-ref">Compute and Reduce with Tuple Inputs</span></a> (<code class="docutils literal notranslate"><span class="pre">tuple_inputs.py</span></code>)</p></li>
+<li><p><strong>00:01.844</strong>: <a class="reference internal" href="intrin_math.html#sphx-glr-how-to-work-with-schedules-intrin-math-py"><span class="std std-ref">Intrinsics and Math Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">intrin_math.py</span></code>)</p></li>
+<li><p><strong>00:00.764</strong>: <a class="reference internal" href="tensorize.html#sphx-glr-how-to-work-with-schedules-tensorize-py"><span class="std std-ref">Use Tensorize to Leverage Hardware Intrinsics</span></a> (<code class="docutils literal notranslate"><span class="pre">tensorize.py</span></code>)</p></li>
+<li><p><strong>00:00.666</strong>: <a class="reference internal" href="reduction.html#sphx-glr-how-to-work-with-schedules-reduction-py"><span class="std std-ref">Reduction</span></a> (<code class="docutils literal notranslate"><span class="pre">reduction.py</span></code>)</p></li>
+<li><p><strong>00:00.657</strong>: <a class="reference internal" href="scan.html#sphx-glr-how-to-work-with-schedules-scan-py"><span class="std std-ref">Scan and Recurrent Kernel</span></a> (<code class="docutils literal notranslate"><span class="pre">scan.py</span></code>)</p></li>
+<li><p><strong>00:00.301</strong>: <a class="reference internal" href="extern_op.html#sphx-glr-how-to-work-with-schedules-extern-op-py"><span class="std std-ref">External Tensor Functions</span></a> (<code class="docutils literal notranslate"><span class="pre">extern_op.py</span></code>)</p></li>
+<li><p><strong>00:00.239</strong>: <a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></li>
+<li><p><strong>00:00.232</strong>: <a class="reference internal" href="tedd.html#sphx-glr-how-to-work-with-schedules-tedd-py"><span class="std std-ref">Use Tensor Expression Debug Display (TEDD) for Visualization</span></a> (<code class="docutils literal notranslate"><span class="pre">tedd.py</span></code>)</p></li>
+<li><p><strong>00:00.219</strong>: <a class="reference internal" href="tuple_inputs.html#sphx-glr-how-to-work-with-schedules-tuple-inputs-py"><span class="std std-ref">Compute and Reduce with Tuple Inputs</span></a> (<code class="docutils literal notranslate"><span class="pre">tuple_inputs.py</span></code>)</p></li>
</ul>
</div>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index 1c12dfa34..6ff7f415b 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -552,7 +552,7 @@ The importing needs to happen before the tensorized GEMV being executed.</p>
C: Buffer(C_2: Pointer(float32), float32, [524288], [])}
buffer_map = {A_1: A, B_1: B, C_1: C}
preflattened_buffer_map = {A_1: A_3: Buffer(A_2, float32, [1024, 64], []), B_1: B_3: Buffer(B_2, float32, [512, 64], []), C_1: C_3: Buffer(C_2, float32, [1024, 512], [])} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpoxih48lj/input0.cc'\nsource_filename = \"/tmp/tmpoxih48lj/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = allo [...]
+ attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmp3nlbt5gp/input0.cc'\nsource_filename = \"/tmp/tmp3nlbt5gp/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n %7 = allo [...]
for (i, 0, 1024) {
for (j.outer: int32, 0, 32) {
@tir.call_extern("gemv_update", @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/reference/api/doxygen/affine__type_8h.html b/docs/reference/api/doxygen/affine__type_8h.html
index f21240728..c978faea1 100644
--- a/docs/reference/api/doxygen/affine__type_8h.html
+++ b/docs/reference/api/doxygen/affine__type_8h.html
@@ -77,7 +77,7 @@ $(function() {
</div><div class="textblock"><div class="dynheader">
Include dependency graph for affine_type.h:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="affine__type_8h__incl.svg" width="3902" height="1082"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
+<div class="center"><iframe scrolling="no" frameborder="0" src="affine__type_8h__incl.svg" width="3942" height="1082"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
</div>
diff --git a/docs/reference/api/doxygen/affine__type_8h__incl.svg b/docs/reference/api/doxygen/affine__type_8h__incl.svg
index 4ebc0a30d..e889d3d85 100644
--- a/docs/reference/api/doxygen/affine__type_8h__incl.svg
+++ b/docs/reference/api/doxygen/affine__type_8h__incl.svg
@@ -4,1289 +4,1301 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: include/tvm/ir/affine_type.h Pages: 1 -->
-<svg width="2926pt" height="811pt"
- viewBox="0.00 0.00 2926.00 811.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<svg width="2956pt" height="811pt"
+ viewBox="0.00 0.00 2956.00 811.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 807)">
<title>include/tvm/ir/affine_type.h</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-807 2922,-807 2922,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-807 2952,-807 2952,4 -4,4"/>
<!-- Node0 -->
<g id="node1" class="node">
<title>Node0</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="329.5,-772.5 329.5,-802.5 444.5,-802.5 444.5,-772.5 329.5,-772.5"/>
-<text text-anchor="start" x="337.5" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/affine</text>
-<text text-anchor="middle" x="387" y="-779.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_type.h</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="2593.5,-772.5 2593.5,-802.5 2708.5,-802.5 2708.5,-772.5 2593.5,-772.5"/>
+<text text-anchor="start" x="2601.5" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/affine</text>
+<text text-anchor="middle" x="2651" y="-779.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_type.h</text>
</g>
<!-- Node1 -->
<g id="node2" class="node">
<title>Node1</title>
<g id="a_node2"><a xlink:href="ir_2expr_8h.html" target="_top" xlink:title="Base expr nodes in TVM. ">
-<polygon fill="#ffffff" stroke="#000000" points="313.5,-716.5 313.5,-735.5 392.5,-735.5 392.5,-716.5 313.5,-716.5"/>
-<text text-anchor="middle" x="353" y="-723.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/expr.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2683.5,-716.5 2683.5,-735.5 2762.5,-735.5 2762.5,-716.5 2683.5,-716.5"/>
+<text text-anchor="middle" x="2723" y="-723.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/expr.h</text>
</a>
</g>
</g>
<!-- Node0->Node1 -->
<g id="edge1" class="edge">
<title>Node0->Node1</title>
-<path fill="none" stroke="#191970" d="M378.5955,-772.2977C373.9743,-763.9388 368.2031,-753.4997 363.3111,-744.6509"/>
-<polygon fill="#191970" stroke="#191970" points="366.3452,-742.9051 358.4438,-735.8469 360.2191,-746.292 366.3452,-742.9051"/>
-</g>
-<!-- Node46 -->
-<g id="node47" class="node">
-<title>Node46</title>
-<g id="a_node47"><a xlink:href="ir_2type_8h.html" target="_top" xlink:title="IR/AST nodes for the unified type system in TVM. ">
-<polygon fill="#ffffff" stroke="#000000" points="382,-660.5 382,-679.5 462,-679.5 462,-660.5 382,-660.5"/>
-<text text-anchor="middle" x="422" y="-667.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/type.h</text>
+<path fill="none" stroke="#191970" d="M2668.7978,-772.2977C2679.4167,-763.2274 2692.9031,-751.7077 2703.7667,-742.4285"/>
+<polygon fill="#191970" stroke="#191970" points="2706.1413,-745.0031 2711.4719,-735.8469 2701.5949,-739.6805 2706.1413,-745.0031"/>
+</g>
+<!-- Node47 -->
+<g id="node48" class="node">
+<title>Node47</title>
+<g id="a_node48"><a xlink:href="ir_2type_8h.html" target="_top" xlink:title="IR/AST nodes for the unified type system in TVM. ">
+<polygon fill="#ffffff" stroke="#000000" points="2512,-660.5 2512,-679.5 2592,-679.5 2592,-660.5 2512,-660.5"/>
+<text text-anchor="middle" x="2552" y="-667.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/type.h</text>
</a>
</g>
</g>
-<!-- Node0->Node46 -->
-<g id="edge153" class="edge">
-<title>Node0->Node46</title>
-<path fill="none" stroke="#191970" d="M391.5756,-772.1389C397.9102,-750.873 409.3493,-712.4702 416.2524,-689.2956"/>
-<polygon fill="#191970" stroke="#191970" points="419.652,-690.1425 419.1525,-679.5595 412.9433,-688.1441 419.652,-690.1425"/>
+<!-- Node0->Node47 -->
+<g id="edge154" class="edge">
+<title>Node0->Node47</title>
+<path fill="none" stroke="#191970" d="M2638.0575,-772.1389C2619.6534,-750.2957 2586.0159,-710.3724 2566.6962,-687.4424"/>
+<polygon fill="#191970" stroke="#191970" points="2569.1744,-684.9517 2560.0544,-679.5595 2563.8212,-689.4621 2569.1744,-684.9517"/>
</g>
<!-- Node2 -->
<g id="node3" class="node">
<title>Node2</title>
<g id="a_node3"><a xlink:href="ir_2span_8h.html" target="_top" xlink:title="Span information for debugging purposes. ">
-<polygon fill="#ffffff" stroke="#000000" points="312.5,-604.5 312.5,-623.5 393.5,-623.5 393.5,-604.5 312.5,-604.5"/>
-<text text-anchor="middle" x="353" y="-611.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/span.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2724.5,-604.5 2724.5,-623.5 2805.5,-623.5 2805.5,-604.5 2724.5,-604.5"/>
+<text text-anchor="middle" x="2765" y="-611.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/span.h</text>
</a>
</g>
</g>
<!-- Node1->Node2 -->
<g id="edge2" class="edge">
<title>Node1->Node2</title>
-<path fill="none" stroke="#191970" d="M353,-716.4509C353,-698.184 353,-657.9553 353,-633.6976"/>
-<polygon fill="#191970" stroke="#191970" points="356.5001,-633.5249 353,-623.5249 349.5001,-633.5249 356.5001,-633.5249"/>
+<path fill="none" stroke="#191970" d="M2726.5809,-716.4509C2733.493,-698.0186 2748.7905,-657.2254 2757.8589,-633.043"/>
+<polygon fill="#191970" stroke="#191970" points="2761.194,-634.1171 2761.4282,-623.5249 2754.6397,-631.6592 2761.194,-634.1171"/>
</g>
<!-- Node3 -->
<g id="node4" class="node">
<title>Node3</title>
<g id="a_node4"><a xlink:href="node_8h.html" target="_top" xlink:title="Definitions and helper macros for IR/AST nodes. ">
-<polygon fill="#ffffff" stroke="#000000" points="398.5,-548.5 398.5,-567.5 497.5,-567.5 497.5,-548.5 398.5,-548.5"/>
-<text text-anchor="middle" x="448" y="-555.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/node.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2218.5,-548.5 2218.5,-567.5 2317.5,-567.5 2317.5,-548.5 2218.5,-548.5"/>
+<text text-anchor="middle" x="2268" y="-555.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/node.h</text>
</a>
</g>
</g>
<!-- Node1->Node3 -->
-<g id="edge146" class="edge">
+<g id="edge147" class="edge">
<title>Node1->Node3</title>
-<path fill="none" stroke="#191970" d="M342.9876,-716.3438C321.6885,-694.6257 275.962,-640.9435 303,-604 313.5147,-589.6331 353.1725,-577.6402 388.36,-569.5056"/>
-<polygon fill="#191970" stroke="#191970" points="389.1454,-572.9165 398.1382,-567.3147 387.6149,-566.0858 389.1454,-572.9165"/>
+<path fill="none" stroke="#191970" d="M2707.678,-716.4784C2685.0086,-702.6792 2640.9051,-676.926 2601,-660 2503.9772,-618.847 2384.7012,-586.4482 2318.453,-569.9605"/>
+<polygon fill="#191970" stroke="#191970" points="2319.2273,-566.5467 2308.6799,-567.5481 2317.5497,-573.3427 2319.2273,-566.5467"/>
</g>
<!-- Node8 -->
<g id="node9" class="node">
<title>Node8</title>
<g id="a_node9"><a xlink:href="object_8h.html" target="_top" xlink:title="A managed object in the TVM runtime. ">
-<polygon fill="#ffffff" stroke="#000000" points="1856.5,-123.5 1856.5,-142.5 1975.5,-142.5 1975.5,-123.5 1856.5,-123.5"/>
-<text text-anchor="middle" x="1916" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/object.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1396.5,-123.5 1396.5,-142.5 1515.5,-142.5 1515.5,-123.5 1396.5,-123.5"/>
+<text text-anchor="middle" x="1456" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/object.h</text>
</a>
</g>
</g>
<!-- Node1->Node8 -->
-<g id="edge148" class="edge">
+<g id="edge149" class="edge">
<title>Node1->Node8</title>
-<path fill="none" stroke="#191970" d="M392.519,-725.3449C716.1332,-719.8554 2918,-680.077 2918,-614 2918,-614 2918,-614 2918,-440.5 2918,-363.6236 2889.7173,-340.6211 2831,-291 2776.9659,-245.3365 2749.1323,-257.318 2682,-235 2597.2453,-206.8235 2576.8525,-195.1003 2489,-179 2394.3797,-161.6594 2117.3317,-144.3498 1986.2651,-136.8579"/>
-<polygon fill="#191970" stroke="#191970" points="1986.127,-133.3445 1975.9444,-136.2711 1985.7296,-140.3332 1986.127,-133.3445"/>
+<path fill="none" stroke="#191970" d="M2758.3109,-716.4663C2802.3722,-702.1751 2872,-670.632 2872,-614 2872,-614 2872,-614 2872,-502 2872,-368.1152 2774.4718,-355.2301 2657,-291 2460.9917,-183.8285 2388.0609,-211.2035 2167,-179 1933.7127,-145.0154 1653.7047,-136.137 1525.8439,-133.8186"/>
+<polygon fill="#191970" stroke="#191970" points="1525.8539,-130.3184 1515.7947,-133.6443 1525.7325,-137.3173 1525.8539,-130.3184"/>
</g>
<!-- Node14 -->
<g id="node15" class="node">
<title>Node14</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="218,-62 218,-81 262,-81 262,-62 218,-62"/>
-<text text-anchor="middle" x="240" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">string</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="2637,-62 2637,-81 2681,-81 2681,-62 2637,-62"/>
+<text text-anchor="middle" x="2659" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">string</text>
</g>
<!-- Node1->Node14 -->
-<g id="edge151" class="edge">
+<g id="edge152" class="edge">
<title>Node1->Node14</title>
-<path fill="none" stroke="#191970" d="M313.1926,-721.8994C220.8741,-711.2884 0,-679.0991 0,-614 0,-614 0,-614 0,-189 0,-151.6561 20.8749,-143.6348 52,-123 100.3152,-90.9689 168.1448,-78.7454 207.7421,-74.1632"/>
-<polygon fill="#191970" stroke="#191970" points="208.3102,-77.6227 217.8847,-73.0862 207.571,-70.6618 208.3102,-77.6227"/>
+<path fill="none" stroke="#191970" d="M2762.7015,-721.4409C2814.2394,-714.7765 2900.3893,-700.8549 2924,-680 2947.3935,-659.3369 2948,-645.2125 2948,-614 2948,-614 2948,-614 2948,-189 2948,-154.6999 2936.0165,-142.7881 2908,-123 2873.3766,-98.5455 2750.5701,-81.7988 2691.3726,-74.9578"/>
+<polygon fill="#191970" stroke="#191970" points="2691.4248,-71.4415 2681.095,-73.796 2690.6385,-78.3972 2691.4248,-71.4415"/>
</g>
<!-- Node15 -->
<g id="node16" class="node">
<title>Node15</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="623.5,-62 623.5,-81 692.5,-81 692.5,-62 623.5,-62"/>
-<text text-anchor="middle" x="658" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">type_traits</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="2263.5,-62 2263.5,-81 2332.5,-81 2332.5,-62 2263.5,-62"/>
+<text text-anchor="middle" x="2298" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">type_traits</text>
</g>
<!-- Node1->Node15 -->
-<g id="edge152" class="edge">
+<g id="edge153" class="edge">
<title>Node1->Node15</title>
-<path fill="none" stroke="#191970" d="M313.3151,-722.9879C251.1921,-715.8711 138,-692.1773 138,-614 138,-614 138,-614 138,-558 138,-367.5406 182.9466,-289.6038 338,-179 378.2129,-150.3151 388.074,-141.1854 434,-123 493.8438,-99.3035 567.2032,-85.0228 613.2882,-77.7001"/>
-<polygon fill="#191970" stroke="#191970" points="613.9118,-81.1453 623.2588,-76.1569 612.8411,-74.2277 613.9118,-81.1453"/>
+<path fill="none" stroke="#191970" d="M2762.6579,-720.2581C2817.55,-709.8156 2910,-682.1171 2910,-614 2910,-614 2910,-614 2910,-502 2910,-224.4556 2483.7241,-110.3232 2342.6978,-80.1587"/>
+<polygon fill="#191970" stroke="#191970" points="2343.1631,-76.6802 2332.6573,-78.0506 2341.7247,-83.5308 2343.1631,-76.6802"/>
</g>
<!-- Node20 -->
<g id="node21" class="node">
<title>Node20</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1628,-179.5 1628,-198.5 1692,-198.5 1692,-179.5 1628,-179.5"/>
-<text text-anchor="middle" x="1660" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">algorithm</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1610,-179.5 1610,-198.5 1674,-198.5 1674,-179.5 1610,-179.5"/>
+<text text-anchor="middle" x="1642" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">algorithm</text>
</g>
<!-- Node1->Node20 -->
-<g id="edge149" class="edge">
+<g id="edge150" class="edge">
<title>Node1->Node20</title>
-<path fill="none" stroke="#191970" d="M392.7254,-725.5913C652.4692,-722.6663 2098,-702.4161 2098,-614 2098,-614 2098,-614 2098,-502 2098,-267.0704 1735.9028,-514.336 1601,-322 1593.0884,-310.7202 1597.136,-304.2249 1601,-291 1606.3248,-272.7753 1614.3372,-271.3439 1624,-255 1633.3161,-239.2424 1643.2296,-220.9412 1650.2898,-207.6128"/>
-<polygon fill="#191970" stroke="#191970" points="1653.5179,-208.9941 1655.0754,-198.5144 1647.3226,-205.7355 1653.5179,-208.9941"/>
+<path fill="none" stroke="#191970" d="M2720.7353,-716.2936C2709.4522,-669.9452 2655.1232,-468.8319 2538,-358 2412.1156,-238.8775 1846.0511,-199.8936 1684.919,-191.1079"/>
+<polygon fill="#191970" stroke="#191970" points="1684.636,-187.5879 1674.4636,-190.5496 1684.2627,-194.578 1684.636,-187.5879"/>
</g>
<!-- Node31 -->
<g id="node32" class="node">
<title>Node31</title>
<g id="a_node32"><a xlink:href="string_8h.html" target="_top" xlink:title="Runtime String container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1170,-291.5 1170,-321.5 1296,-321.5 1296,-291.5 1170,-291.5"/>
-<text text-anchor="start" x="1178" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
-<text text-anchor="middle" x="1233" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/string.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1274,-291.5 1274,-321.5 1400,-321.5 1400,-291.5 1274,-291.5"/>
+<text text-anchor="start" x="1282" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
+<text text-anchor="middle" x="1337" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/string.h</text>
</a>
</g>
</g>
<!-- Node1->Node31 -->
-<g id="edge147" class="edge">
+<g id="edge148" class="edge">
<title>Node1->Node31</title>
-<path fill="none" stroke="#191970" d="M392.7245,-723.2691C544.8695,-712.5396 1090.3028,-671.525 1257,-624 1312.8262,-608.0841 1343.5385,-616.7858 1375,-568 1430.1269,-482.5173 1316.4587,-373.8403 1261.0306,-328.2128"/>
-<polygon fill="#191970" stroke="#191970" points="1262.9591,-325.2708 1252.9855,-321.6961 1258.5531,-330.7102 1262.9591,-325.2708"/>
+<path fill="none" stroke="#191970" d="M2683.2474,-725.4832C2429.5951,-721.9321 1048,-698.7021 1048,-614 1048,-614 1048,-614 1048,-440.5 1048,-392.4398 1181.2068,-347.8733 1266.7486,-324.2087"/>
+<polygon fill="#191970" stroke="#191970" points="1267.6925,-327.5792 1276.4161,-321.5668 1265.8472,-320.8268 1267.6925,-327.5792"/>
</g>
-<!-- Node42 -->
-<g id="node43" class="node">
-<title>Node42</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="332,-364 332,-383 376,-383 376,-364 332,-364"/>
-<text text-anchor="middle" x="354" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">limits</text>
+<!-- Node43 -->
+<g id="node44" class="node">
+<title>Node43</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="2458,-364 2458,-383 2502,-383 2502,-364 2458,-364"/>
+<text text-anchor="middle" x="2480" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">limits</text>
</g>
-<!-- Node1->Node42 -->
-<g id="edge150" class="edge">
-<title>Node1->Node42</title>
-<path fill="none" stroke="#191970" d="M336.1298,-716.4189C323.6774,-708.3859 307.4795,-695.7156 299,-680 287.0896,-657.9256 293.5586,-649.0341 292,-624 286.4823,-535.3729 326.6734,-433.8234 345.1025,-392.4874"/>
-<polygon fill="#191970" stroke="#191970" points="348.318,-393.8713 349.2779,-383.3201 341.9477,-390.9698 348.318,-393.8713"/>
+<!-- Node1->Node43 -->
+<g id="edge151" class="edge">
+<title>Node1->Node43</title>
+<path fill="none" stroke="#191970" d="M2740.0302,-716.4704C2772.7539,-696.8464 2839.1356,-649.9036 2815,-604 2749.3963,-479.2278 2582.3208,-408.8641 2511.9825,-383.943"/>
+<polygon fill="#191970" stroke="#191970" points="2512.925,-380.5654 2502.3302,-380.5948 2510.6309,-387.1788 2512.925,-380.5654"/>
</g>
-<!-- Node1->Node46 -->
-<g id="edge139" class="edge">
-<title>Node1->Node46</title>
-<path fill="none" stroke="#191970" d="M365.0189,-716.2455C375.2098,-707.9746 390.0362,-695.9416 402.0107,-686.2232"/>
-<polygon fill="#191970" stroke="#191970" points="404.5598,-688.662 410.1188,-679.6427 400.1486,-683.2268 404.5598,-688.662"/>
+<!-- Node1->Node47 -->
+<g id="edge140" class="edge">
+<title>Node1->Node47</title>
+<path fill="none" stroke="#191970" d="M2693.6041,-716.3733C2665.1693,-707.0613 2621.9123,-692.8953 2590.6919,-682.671"/>
+<polygon fill="#191970" stroke="#191970" points="2591.605,-679.2872 2581.0124,-679.5011 2589.4264,-685.9396 2591.605,-679.2872"/>
</g>
<!-- Node2->Node3 -->
<g id="edge3" class="edge">
<title>Node2->Node3</title>
-<path fill="none" stroke="#191970" d="M369.5477,-604.2455C384.2599,-595.5731 405.9888,-582.7645 422.8505,-572.825"/>
-<polygon fill="#191970" stroke="#191970" points="424.8046,-575.736 431.6419,-567.6427 421.2498,-569.7057 424.8046,-575.736"/>
+<path fill="none" stroke="#191970" d="M2724.3284,-609.4173C2636.8038,-599.5554 2430.2421,-576.2808 2327.781,-564.7359"/>
+<polygon fill="#191970" stroke="#191970" points="2327.9309,-561.2307 2317.6018,-563.5889 2327.147,-568.1867 2327.9309,-561.2307"/>
</g>
<!-- Node2->Node8 -->
-<g id="edge137" class="edge">
+<g id="edge138" class="edge">
<title>Node2->Node8</title>
-<path fill="none" stroke="#191970" d="M393.6449,-613.1146C696.394,-606.4108 2591.0903,-562.53 2700,-512 2739.2299,-493.7988 2766,-483.7466 2766,-440.5 2766,-440.5 2766,-440.5 2766,-373.5 2766,-273.6124 2683.5925,-267.0903 2589,-235 2377.0145,-163.0844 2110.2667,-141.7491 1985.8629,-135.516"/>
-<polygon fill="#191970" stroke="#191970" points="1985.8703,-132.0123 1975.7129,-135.0253 1985.5322,-139.0041 1985.8703,-132.0123"/>
+<path fill="none" stroke="#191970" d="M2764.5407,-604.2909C2762.4305,-568.514 2751.2237,-441.7841 2696,-358 2671.2189,-320.4028 2660.3911,-310.9051 2620,-291 2545.2551,-254.1651 2517.5515,-272.116 2436,-255 2284.8858,-223.2842 2249.7618,-201.4773 2097,-179 1890.739,-148.6509 1644.0703,-138.0986 1525.8764,-134.6124"/>
+<polygon fill="#191970" stroke="#191970" points="1525.6267,-131.1039 1515.5304,-134.3157 1525.4259,-138.101 1525.6267,-131.1039"/>
</g>
<!-- Node2->Node14 -->
-<g id="edge138" class="edge">
+<g id="edge139" class="edge">
<title>Node2->Node14</title>
-<path fill="none" stroke="#191970" d="M323.6184,-604.4368C252.3291,-579.1856 76,-503.5806 76,-373.5 76,-373.5 76,-373.5 76,-189 76,-124.7162 160.2933,-92.2129 208.0157,-78.9623"/>
-<polygon fill="#191970" stroke="#191970" points="209.1103,-82.2932 217.8822,-76.3515 207.3196,-75.5261 209.1103,-82.2932"/>
+<path fill="none" stroke="#191970" d="M2773.6362,-604.1906C2804.852,-567.6852 2910,-435.1701 2910,-306.5 2910,-306.5 2910,-306.5 2910,-189 2910,-157.7875 2909.413,-143.6411 2886,-123 2857.3087,-97.7055 2746.9278,-81.6595 2691.3566,-75.0151"/>
+<polygon fill="#191970" stroke="#191970" points="2691.6693,-71.528 2681.3316,-73.8476 2690.8595,-78.481 2691.6693,-71.528"/>
</g>
<!-- Node4 -->
<g id="node5" class="node">
<title>Node4</title>
<g id="a_node5"><a xlink:href="reflection_8h.html" target="_top" xlink:title="Reflection and serialization of compiler IR/AST nodes. ">
-<polygon fill="#ffffff" stroke="#000000" points="575.5,-492.5 575.5,-511.5 696.5,-511.5 696.5,-492.5 575.5,-492.5"/>
-<text text-anchor="middle" x="636" y="-499.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/reflection.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1831.5,-492.5 1831.5,-511.5 1952.5,-511.5 1952.5,-492.5 1831.5,-492.5"/>
+<text text-anchor="middle" x="1892" y="-499.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/reflection.h</text>
</a>
</g>
</g>
<!-- Node3->Node4 -->
<g id="edge4" class="edge">
<title>Node3->Node4</title>
-<path fill="none" stroke="#191970" d="M480.3184,-548.3733C511.8506,-538.9807 559.9621,-524.6496 594.3501,-514.4064"/>
-<polygon fill="#191970" stroke="#191970" points="595.5187,-517.7103 604.1034,-511.5011 593.5203,-511.0016 595.5187,-517.7103"/>
+<path fill="none" stroke="#191970" d="M2218.4579,-550.6214C2153.0361,-540.8777 2037.4186,-523.6581 1962.7927,-512.5436"/>
+<polygon fill="#191970" stroke="#191970" points="1962.9508,-509.0286 1952.5442,-511.0172 1961.9195,-515.9522 1962.9508,-509.0286"/>
</g>
<!-- Node5 -->
<g id="node6" class="node">
<title>Node5</title>
<g id="a_node6"><a xlink:href="structural__equal_8h.html" target="_top" xlink:title="Structural equality comparison. ">
-<polygon fill="#ffffff" stroke="#000000" points="622.5,-358.5 622.5,-388.5 735.5,-388.5 735.5,-358.5 622.5,-358.5"/>
-<text text-anchor="start" x="630.5" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/structural</text>
-<text text-anchor="middle" x="679" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_equal.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2136.5,-358.5 2136.5,-388.5 2249.5,-388.5 2249.5,-358.5 2136.5,-358.5"/>
+<text text-anchor="start" x="2144.5" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/structural</text>
+<text text-anchor="middle" x="2193" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_equal.h</text>
</a>
</g>
</g>
<!-- Node3->Node5 -->
-<g id="edge128" class="edge">
+<g id="edge129" class="edge">
<title>Node3->Node5</title>
-<path fill="none" stroke="#191970" d="M460.0943,-548.3402C495.8845,-519.7546 601.8967,-435.0825 652.0828,-394.9988"/>
-<polygon fill="#191970" stroke="#191970" points="654.2942,-397.7119 659.9236,-388.7363 649.9257,-392.2424 654.2942,-397.7119"/>
+<path fill="none" stroke="#191970" d="M2261.6289,-548.4312C2255.7534,-539.3296 2247.0509,-525.1175 2241,-512 2223.325,-473.6827 2208.4089,-426.8967 2200.0931,-398.6884"/>
+<polygon fill="#191970" stroke="#191970" points="2203.3847,-397.4722 2197.2372,-388.8433 2196.6619,-399.4225 2203.3847,-397.4722"/>
</g>
<!-- Node3->Node8 -->
-<g id="edge132" class="edge">
+<g id="edge133" class="edge">
<title>Node3->Node8</title>
-<path fill="none" stroke="#191970" d="M497.5975,-557.5429C821.0392,-554.4929 2618.4608,-536.4631 2669,-512 2706.084,-494.0498 2728,-481.6999 2728,-440.5 2728,-440.5 2728,-440.5 2728,-373.5 2728,-325.6995 2696.4979,-320.6445 2659,-291 2615.3419,-256.4855 2600.1264,-251.5793 2547,-235 2438.5798,-201.165 2405.9142,-218.3641 2294,-199 2250.1975,-191.421 2239.7185,-187.0492 2196,-179 2124.4728,-165.8309 2042.3215,-152.5522 1985.5888,-143.6701"/>
-<polygon fill="#191970" stroke="#191970" points="1986.0729,-140.2034 1975.6526,-142.1188 1984.993,-147.1196 1986.0729,-140.2034"/>
+<path fill="none" stroke="#191970" d="M2317.6186,-549.7812C2370.2155,-538.3267 2451.5573,-512.2503 2495,-456 2521.9754,-421.0718 2532.918,-396.3048 2511,-358 2411.7744,-184.5895 1745.1034,-143.9432 1525.7881,-135.2141"/>
+<polygon fill="#191970" stroke="#191970" points="1525.815,-131.7126 1515.6867,-134.8222 1525.5435,-138.7074 1525.815,-131.7126"/>
</g>
<!-- Node9 -->
<g id="node10" class="node">
<title>Node9</title>
<g id="a_node10"><a xlink:href="c__runtime__api_8h.html" target="_top" xlink:title="tvm/runtime/c_runtime\l_api.h">
-<polygon fill="#ffffff" stroke="#000000" points="782.5,-56.5 782.5,-86.5 911.5,-86.5 911.5,-56.5 782.5,-56.5"/>
-<text text-anchor="start" x="790.5" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/c_runtime</text>
-<text text-anchor="middle" x="847" y="-63.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_api.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="125.5,-56.5 125.5,-86.5 254.5,-86.5 254.5,-56.5 125.5,-56.5"/>
+<text text-anchor="start" x="133.5" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/c_runtime</text>
+<text text-anchor="middle" x="190" y="-63.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_api.h</text>
</a>
</g>
</g>
<!-- Node3->Node9 -->
-<g id="edge130" class="edge">
+<g id="edge131" class="edge">
<title>Node3->Node9</title>
-<path fill="none" stroke="#191970" d="M433.5767,-548.4247C392.5833,-519.9777 280.9941,-433.2628 323,-358 417.119,-189.3651 651.6783,-114.4547 772.4855,-86.1561"/>
-<polygon fill="#191970" stroke="#191970" points="773.4361,-89.529 782.3987,-83.8789 771.8689,-82.7066 773.4361,-89.529"/>
+<path fill="none" stroke="#191970" d="M2218.4065,-557.5653C1904.6696,-554.7365 207.1939,-538.237 161,-512 40.5077,-443.5632 0,-383.5712 0,-245 0,-245 0,-245 0,-189 0,-132.133 62.3281,-101.8622 115.5875,-86.3764"/>
+<polygon fill="#191970" stroke="#191970" points="116.7078,-89.6979 125.4095,-83.6538 114.838,-82.9522 116.7078,-89.6979"/>
</g>
<!-- Node3->Node14 -->
-<g id="edge133" class="edge">
+<g id="edge134" class="edge">
<title>Node3->Node14</title>
-<path fill="none" stroke="#191970" d="M398.3575,-553.3452C305.3154,-541.6017 114,-501.8495 114,-373.5 114,-373.5 114,-373.5 114,-189 114,-137.518 170.9572,-102.148 208.2824,-84.4965"/>
-<polygon fill="#191970" stroke="#191970" points="210.1064,-87.5122 217.7638,-80.19 207.2115,-81.1388 210.1064,-87.5122"/>
+<path fill="none" stroke="#191970" d="M2317.6156,-554.0357C2363.558,-548.9048 2432.8894,-537.3868 2488,-512 2576.211,-471.3655 2834,-342.1203 2834,-245 2834,-245 2834,-245 2834,-189 2834,-120.8697 2741.446,-89.6903 2690.9562,-77.7502"/>
+<polygon fill="#191970" stroke="#191970" points="2691.6879,-74.3274 2681.1639,-75.5505 2690.1536,-81.1572 2691.6879,-74.3274"/>
</g>
<!-- Node3->Node15 -->
-<g id="edge134" class="edge">
+<g id="edge135" class="edge">
<title>Node3->Node15</title>
-<path fill="none" stroke="#191970" d="M415.8867,-548.4978C344.3287,-525.3306 180.6044,-460.2354 214,-358 271.2822,-182.6392 512.5397,-105.9945 613.3933,-81.1841"/>
-<polygon fill="#191970" stroke="#191970" points="614.3691,-84.5494 623.2744,-78.8094 612.7333,-77.7432 614.3691,-84.5494"/>
+<path fill="none" stroke="#191970" d="M2317.7898,-552.414C2357.0677,-546.4883 2412.3227,-534.7055 2456,-512 2489.5354,-494.5667 2501.6364,-489.0352 2520,-456 2534.6615,-429.6247 2528.5474,-419.0766 2531,-389 2542.2878,-250.5736 2463.6292,-222.7601 2367,-123 2353.606,-109.172 2336.2252,-96.223 2322.2589,-86.7684"/>
+<polygon fill="#191970" stroke="#191970" points="2323.9243,-83.6739 2313.6504,-81.0858 2320.0679,-89.5159 2323.9243,-83.6739"/>
</g>
<!-- Node16 -->
<g id="node17" class="node">
<title>Node16</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2380.5,-62 2380.5,-81 2425.5,-81 2425.5,-62 2380.5,-62"/>
-<text text-anchor="middle" x="2403" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">utility</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="859.5,-62 859.5,-81 904.5,-81 904.5,-62 859.5,-62"/>
+<text text-anchor="middle" x="882" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">utility</text>
</g>
<!-- Node3->Node16 -->
-<g id="edge135" class="edge">
+<g id="edge136" class="edge">
<title>Node3->Node16</title>
-<path fill="none" stroke="#191970" d="M497.5014,-557.5597C823.3157,-554.5929 2648.7414,-536.8801 2700,-512 2837.8196,-445.1046 2918,-398.1967 2918,-245 2918,-245 2918,-245 2918,-189 2918,-89.7197 2549.6117,-74.3251 2435.9744,-71.938"/>
-<polygon fill="#191970" stroke="#191970" points="2435.8579,-68.4353 2425.7934,-71.7453 2435.7253,-75.434 2435.8579,-68.4353"/>
+<path fill="none" stroke="#191970" d="M2218.166,-557.3046C1946.3371,-553.4341 648.3158,-533.9283 471,-512 326.803,-494.1675 152,-585.7954 152,-440.5 152,-440.5 152,-440.5 152,-373.5 152,-219.6225 245.453,-172.9434 391,-123 476.3804,-93.7024 753.6546,-77.7519 849.2488,-73.0186"/>
+<polygon fill="#191970" stroke="#191970" points="849.6523,-76.5032 859.4703,-72.5212 849.312,-69.5115 849.6523,-76.5032"/>
</g>
<!-- Node18 -->
<g id="node19" class="node">
<title>Node18</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1962.5,-235.5 1962.5,-254.5 2009.5,-254.5 2009.5,-235.5 1962.5,-235.5"/>
-<text text-anchor="middle" x="1986" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">vector</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1652.5,-235.5 1652.5,-254.5 1699.5,-254.5 1699.5,-235.5 1652.5,-235.5"/>
+<text text-anchor="middle" x="1676" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">vector</text>
</g>
<!-- Node3->Node18 -->
-<g id="edge136" class="edge">
+<g id="edge137" class="edge">
<title>Node3->Node18</title>
-<path fill="none" stroke="#191970" d="M497.567,-557.0442C768.6913,-551.4984 2060,-520.8666 2060,-440.5 2060,-440.5 2060,-440.5 2060,-373.5 2060,-335.8735 2060.677,-323.0713 2041,-291 2033.5005,-278.7767 2021.6127,-268.2392 2010.8048,-260.3455"/>
-<polygon fill="#191970" stroke="#191970" points="2012.6925,-257.3957 2002.4731,-254.6001 2008.7186,-263.1584 2012.6925,-257.3957"/>
+<path fill="none" stroke="#191970" d="M2218.1641,-554.722C2118.078,-547.8336 1895.9007,-530.921 1822,-512 1759.867,-496.0919 1742.9783,-490.6399 1689,-456 1651.9372,-432.2154 1633.72,-429.3157 1616,-389 1596.0503,-343.6114 1635.3132,-289.502 1659.3056,-262.3535"/>
+<polygon fill="#191970" stroke="#191970" points="1662.0159,-264.576 1666.1788,-254.8332 1656.8489,-259.8535 1662.0159,-264.576"/>
</g>
<!-- Node23 -->
<g id="node24" class="node">
<title>Node23</title>
<g id="a_node24"><a xlink:href="runtime_2memory_8h.html" target="_top" xlink:title="Runtime memory management. ">
-<polygon fill="#ffffff" stroke="#000000" points="762.5,-179.5 762.5,-198.5 891.5,-198.5 891.5,-179.5 762.5,-179.5"/>
-<text text-anchor="middle" x="827" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/memory.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="948.5,-179.5 948.5,-198.5 1077.5,-198.5 1077.5,-179.5 948.5,-179.5"/>
+<text text-anchor="middle" x="1013" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/memory.h</text>
</a>
</g>
</g>
<!-- Node3->Node23 -->
-<g id="edge131" class="edge">
+<g id="edge132" class="edge">
<title>Node3->Node23</title>
-<path fill="none" stroke="#191970" d="M450.7278,-548.3955C464.2812,-502.7284 529.1351,-306.7993 663,-235 693.0985,-218.8565 729.421,-207.9164 760.2246,-200.7706"/>
-<polygon fill="#191970" stroke="#191970" points="761.0211,-204.179 770.0169,-198.582 759.4943,-197.3475 761.0211,-204.179"/>
+<path fill="none" stroke="#191970" d="M2218.2626,-557.7563C1909.3349,-555.9253 266,-541.4414 266,-440.5 266,-440.5 266,-440.5 266,-373.5 266,-285.6581 327.0128,-271.3077 407,-235 454.2527,-213.5511 786.7564,-197.9026 938.189,-191.8063"/>
+<polygon fill="#191970" stroke="#191970" points="938.4826,-195.2975 948.335,-191.4013 938.2033,-188.303 938.4826,-195.2975"/>
</g>
<!-- Node27 -->
<g id="node28" class="node">
<title>Node27</title>
<g id="a_node28"><a xlink:href="structural__hash_8h.html" target="_top" xlink:title="tvm/node/structural\l_hash.h">
-<polygon fill="#ffffff" stroke="#000000" points="1220.5,-425.5 1220.5,-455.5 1333.5,-455.5 1333.5,-425.5 1220.5,-425.5"/>
-<text text-anchor="start" x="1228.5" y="-443.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/structural</text>
-<text text-anchor="middle" x="1277" y="-432.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_hash.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1937.5,-425.5 1937.5,-455.5 2050.5,-455.5 2050.5,-425.5 1937.5,-425.5"/>
+<text text-anchor="start" x="1945.5" y="-443.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/structural</text>
+<text text-anchor="middle" x="1994" y="-432.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_hash.h</text>
</a>
</g>
</g>
<!-- Node3->Node27 -->
-<g id="edge129" class="edge">
+<g id="edge130" class="edge">
<title>Node3->Node27</title>
-<path fill="none" stroke="#191970" d="M497.8806,-552.7069C619.4515,-539.4983 940.5363,-502.7582 1206,-456 1207.293,-455.7723 1208.5982,-455.5374 1209.9125,-455.2964"/>
-<polygon fill="#191970" stroke="#191970" points="1210.9502,-458.6614 1220.1142,-453.3444 1209.6346,-451.7862 1210.9502,-458.6614"/>
+<path fill="none" stroke="#191970" d="M2245.5774,-548.3845C2200.418,-529.0187 2097.7774,-485.0031 2038.4975,-459.5819"/>
+<polygon fill="#191970" stroke="#191970" points="2039.5877,-456.2413 2029.0176,-455.5167 2036.8288,-462.6747 2039.5877,-456.2413"/>
</g>
-<!-- Node44 -->
-<g id="node45" class="node">
-<title>Node44</title>
-<g id="a_node45"><a xlink:href="repr__printer_8h.html" target="_top" xlink:title="Printer class to print repr string of each AST/IR nodes. ">
-<polygon fill="#ffffff" stroke="#000000" points="2528.5,-492.5 2528.5,-511.5 2659.5,-511.5 2659.5,-492.5 2528.5,-492.5"/>
-<text text-anchor="middle" x="2594" y="-499.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/repr_printer.h</text>
+<!-- Node45 -->
+<g id="node46" class="node">
+<title>Node45</title>
+<g id="a_node46"><a xlink:href="repr__printer_8h.html" target="_top" xlink:title="Printer class to print repr string of each AST/IR nodes. ">
+<polygon fill="#ffffff" stroke="#000000" points="2250.5,-492.5 2250.5,-511.5 2381.5,-511.5 2381.5,-492.5 2250.5,-492.5"/>
+<text text-anchor="middle" x="2316" y="-499.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/repr_printer.h</text>
</a>
</g>
</g>
-<!-- Node3->Node44 -->
-<g id="edge125" class="edge">
-<title>Node3->Node44</title>
-<path fill="none" stroke="#191970" d="M497.6283,-556.7049C779.2496,-549.356 2175.8938,-512.9105 2518.1783,-503.9786"/>
-<polygon fill="#191970" stroke="#191970" points="2518.3299,-507.4759 2528.2351,-503.7161 2518.1472,-500.4783 2518.3299,-507.4759"/>
+<!-- Node3->Node45 -->
+<g id="edge126" class="edge">
+<title>Node3->Node45</title>
+<path fill="none" stroke="#191970" d="M2276.361,-548.2455C2283.175,-540.2958 2292.9681,-528.8706 2301.1152,-519.3656"/>
+<polygon fill="#191970" stroke="#191970" points="2303.8843,-521.513 2307.7349,-511.6427 2298.5695,-516.9575 2303.8843,-521.513"/>
</g>
<!-- Node4->Node5 -->
<g id="edge5" class="edge">
<title>Node4->Node5</title>
-<path fill="none" stroke="#191970" d="M639.232,-492.3416C645.8422,-472.5879 661.0896,-427.0229 670.6674,-398.4008"/>
-<polygon fill="#191970" stroke="#191970" points="674.0588,-399.2954 673.9131,-388.7016 667.4206,-397.0741 674.0588,-399.2954"/>
+<path fill="none" stroke="#191970" d="M1940.3768,-492.4884C1974.398,-484.8905 2020.5307,-472.7347 2059,-456 2097.4877,-439.2572 2138.0409,-412.8974 2164.319,-394.4934"/>
+<polygon fill="#191970" stroke="#191970" points="2166.5246,-397.2201 2172.6598,-388.5824 2162.4772,-391.5089 2166.5246,-397.2201"/>
</g>
<!-- Node4->Node8 -->
<g id="edge89" class="edge">
<title>Node4->Node8</title>
-<path fill="none" stroke="#191970" d="M696.7477,-501.4266C1017.4293,-497.7686 2503.6215,-472.7264 2633,-322 2709.6496,-232.703 2665.9305,-281.1412 2481,-235 2366.5304,-206.4392 2333.5446,-222.8424 2218,-199 2185.4736,-192.2883 2178.3052,-186.7063 2146,-179 2090.0931,-165.6636 2025.9189,-153.05 1979.5881,-144.4162"/>
-<polygon fill="#191970" stroke="#191970" points="1980.1584,-140.9624 1969.688,-142.5813 1978.8827,-147.8452 1980.1584,-140.9624"/>
+<path fill="none" stroke="#191970" d="M1831.4419,-500.8188C1702.048,-497.7533 1389.1999,-487.4522 1129,-456 1052.7535,-446.7836 1034.668,-438.146 959,-425 822.5898,-401.3011 721.977,-448.1864 665,-322 659.3301,-309.443 656.7031,-301.9995 665,-291 719.5571,-218.6721 772.5567,-254.6363 861,-235 959.8645,-213.05 986.8774,-219.7534 1086,-199 1120.6074,-191.7542 1128.382,-186.1951 1163,-179 1238.8934,-163.2261 1326.8489,-150.097 1386.4361,-141.9567"/>
+<polygon fill="#191970" stroke="#191970" points="1387.0458,-145.4062 1396.485,-140.5945 1386.1055,-138.4696 1387.0458,-145.4062"/>
</g>
<!-- Node4->Node9 -->
<g id="edge85" class="edge">
<title>Node4->Node9</title>
-<path fill="none" stroke="#191970" d="M621.1343,-492.2024C609.9816,-483.9678 595.4115,-471.1343 588,-456 555.6882,-390.0191 569.1507,-358.5666 598,-291 635.7231,-202.6506 656.5072,-179.7759 734,-123 752.0094,-109.8052 773.9389,-98.8584 793.6701,-90.5"/>
-<polygon fill="#191970" stroke="#191970" points="795.2319,-93.6423 803.153,-86.6061 792.573,-87.1669 795.2319,-93.6423"/>
+<path fill="none" stroke="#191970" d="M1831.4281,-500.8476C1667.3913,-497.4169 1200.8618,-485.606 814,-456 543.8156,-435.3231 460.1413,-488.2568 208,-389 115.8635,-352.73 38,-344.0184 38,-245 38,-245 38,-245 38,-189 38,-139.8507 88.3612,-108.3523 131.1071,-90.4691"/>
+<polygon fill="#191970" stroke="#191970" points="132.6704,-93.6132 140.65,-86.6434 130.0657,-87.1158 132.6704,-93.6132"/>
</g>
<!-- Node4->Node14 -->
-<g id="edge122" class="edge">
+<g id="edge123" class="edge">
<title>Node4->Node14</title>
-<path fill="none" stroke="#191970" d="M582.8466,-492.4361C545.0674,-484.7807 493.5199,-472.5775 450,-456 390.3625,-433.283 362.6559,-439.001 323,-389 249.4478,-296.26 240.6132,-144.9377 239.8956,-91.1969"/>
-<polygon fill="#191970" stroke="#191970" points="243.3947,-90.9879 239.8464,-81.005 236.3948,-91.0218 243.3947,-90.9879"/>
+<path fill="none" stroke="#191970" d="M1952.5494,-496.4212C2143.4267,-477.1839 2720,-405.7691 2720,-245 2720,-245 2720,-245 2720,-189 2720,-149.8384 2692.979,-110.7568 2674.9051,-88.996"/>
+<polygon fill="#191970" stroke="#191970" points="2677.4426,-86.579 2668.2646,-81.2861 2672.1387,-91.1472 2677.4426,-86.579"/>
</g>
<!-- Node4->Node15 -->
-<g id="edge123" class="edge">
+<g id="edge124" class="edge">
<title>Node4->Node15</title>
-<path fill="none" stroke="#191970" d="M596.1479,-492.4615C569.0125,-484.9464 532.9137,-472.8997 504,-456 429.5674,-412.4952 366.3749,-369.5095 402,-291 448.8808,-187.6855 571.4813,-115.0063 628.2994,-85.7838"/>
-<polygon fill="#191970" stroke="#191970" points="630.1242,-88.7833 637.4725,-81.1509 626.9684,-82.535 630.1242,-88.7833"/>
+<path fill="none" stroke="#191970" d="M1952.5869,-499.4888C2012.7297,-495.3803 2106.7139,-484.5497 2183,-456 2288.0995,-416.667 2342.6412,-420.1739 2397,-322 2442.8702,-239.1569 2354.2274,-130.6349 2315.1742,-88.8942"/>
+<polygon fill="#191970" stroke="#191970" points="2317.5166,-86.2792 2308.0793,-81.4639 2312.4539,-91.1134 2317.5166,-86.2792"/>
</g>
<!-- Node4->Node18 -->
-<g id="edge124" class="edge">
+<g id="edge125" class="edge">
<title>Node4->Node18</title>
-<path fill="none" stroke="#191970" d="M696.5987,-498.6268C820.7973,-491.3933 1095.8805,-473.6126 1135,-456 1153.3491,-447.7388 1149.5702,-433.0795 1168,-425 1269.6192,-380.4507 1558.3549,-415.9007 1666,-389 1768.6663,-363.3435 1786.791,-335.0732 1883,-291 1907.1319,-279.9452 1934.5823,-267.7021 1955.0731,-258.6277"/>
-<polygon fill="#191970" stroke="#191970" points="1956.514,-261.8175 1964.244,-254.5721 1953.6828,-255.4156 1956.514,-261.8175"/>
+<path fill="none" stroke="#191970" d="M1885.2667,-492.2652C1867.8006,-467.5183 1818.7135,-401.114 1766,-358 1741.0371,-337.583 1723.6382,-346.7803 1703,-322 1689.2993,-305.5496 1682.3653,-281.6877 1678.9833,-264.8977"/>
+<polygon fill="#191970" stroke="#191970" points="1682.3484,-263.8246 1677.1699,-254.5816 1675.4541,-265.0366 1682.3484,-263.8246"/>
</g>
<!-- Node4->Node23 -->
<g id="edge87" class="edge">
<title>Node4->Node23</title>
-<path fill="none" stroke="#191970" d="M626.1695,-492.4557C617.9974,-483.8069 606.9367,-470.2647 602,-456 577.7848,-386.0293 575.6143,-346.1663 625,-291 666.8678,-244.2316 735.1598,-216.1095 780.583,-201.5763"/>
-<polygon fill="#191970" stroke="#191970" points="781.6793,-204.9009 790.193,-198.5947 779.605,-198.2153 781.6793,-204.9009"/>
+<path fill="none" stroke="#191970" d="M1831.4051,-500.5716C1695.3032,-496.9322 1355.1098,-485.4285 1072,-456 875.2504,-435.5484 825.232,-431.3096 632,-389 526.7417,-365.9529 463.8146,-408.0787 399,-322 373.6195,-288.2928 400.5338,-256.2265 437,-235 479.3075,-210.3733 791.9564,-196.5584 938.1008,-191.3849"/>
+<polygon fill="#191970" stroke="#191970" points="938.4579,-194.8746 948.3295,-191.0273 938.2132,-187.8789 938.4579,-194.8746"/>
</g>
<!-- Node26 -->
<g id="node27" class="node">
<title>Node26</title>
<g id="a_node27"><a xlink:href="data__type_8h.html" target="_top" xlink:title="tvm/runtime/data_type.h">
-<polygon fill="#ffffff" stroke="#000000" points="634,-297 634,-316 772,-316 772,-297 634,-297"/>
-<text text-anchor="middle" x="703" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/data_type.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2250,-297 2250,-316 2388,-316 2388,-297 2250,-297"/>
+<text text-anchor="middle" x="2319" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/data_type.h</text>
</a>
</g>
</g>
<!-- Node4->Node26 -->
<g id="edge86" class="edge">
<title>Node4->Node26</title>
-<path fill="none" stroke="#191970" d="M630.6447,-492.4792C617.7878,-468.2345 588.0996,-402.9114 614,-358 624.1586,-340.385 642.8971,-328.2497 660.5691,-320.205"/>
-<polygon fill="#191970" stroke="#191970" points="662.3549,-323.2509 670.216,-316.1478 659.6411,-316.7984 662.3549,-323.2509"/>
+<path fill="none" stroke="#191970" d="M1952.6792,-496.76C2030.8089,-487.2294 2167.2945,-460.8458 2258,-389 2280.3703,-371.281 2298.6313,-343.4294 2309.1654,-325.0667"/>
+<polygon fill="#191970" stroke="#191970" points="2312.293,-326.6434 2314.0752,-316.1995 2306.1691,-323.2525 2312.293,-326.6434"/>
</g>
<!-- Node4->Node27 -->
<g id="edge44" class="edge">
<title>Node4->Node27</title>
-<path fill="none" stroke="#191970" d="M696.6261,-499.7851C801.1067,-495.3717 1021.6321,-483.437 1206,-456 1207.2986,-455.8067 1208.609,-455.6036 1209.928,-455.3917"/>
-<polygon fill="#191970" stroke="#191970" points="1210.905,-458.7744 1220.158,-453.6136 1209.7063,-451.8778 1210.905,-458.7744"/>
+<path fill="none" stroke="#191970" d="M1907.9375,-492.3906C1921.8698,-483.9903 1942.5684,-471.5103 1960.1749,-460.8945"/>
+<polygon fill="#191970" stroke="#191970" points="1962.2914,-463.7054 1969.048,-455.5446 1958.677,-457.7108 1962.2914,-463.7054"/>
</g>
<!-- Node28 -->
<g id="node29" class="node">
<title>Node28</title>
<g id="a_node29"><a xlink:href="ndarray_8h.html" target="_top" xlink:title="A device-independent managed NDArray abstraction. ">
-<polygon fill="#ffffff" stroke="#000000" points="1830.5,-364 1830.5,-383 1955.5,-383 1955.5,-364 1830.5,-364"/>
-<text text-anchor="middle" x="1893" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/ndarray.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="750.5,-364 750.5,-383 875.5,-383 875.5,-364 750.5,-364"/>
+<text text-anchor="middle" x="813" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/ndarray.h</text>
</a>
</g>
</g>
<!-- Node4->Node28 -->
<g id="edge88" class="edge">
<title>Node4->Node28</title>
-<path fill="none" stroke="#191970" d="M696.9346,-498.5262C827.2389,-490.8389 1124.6753,-471.8153 1168,-456 1190.4713,-447.7971 1189.4211,-432.902 1212,-425 1338.9126,-380.5839 1682.6583,-406.3091 1816,-389 1824.1718,-387.9392 1832.7761,-386.5156 1841.1535,-384.9595"/>
-<polygon fill="#191970" stroke="#191970" points="1842.0629,-388.3485 1851.2151,-383.0111 1840.7321,-381.4761 1842.0629,-388.3485"/>
+<path fill="none" stroke="#191970" d="M1831.2679,-497.3231C1674.1337,-484.7602 1241.5787,-447.2378 885,-389 878.2172,-387.8922 871.1086,-386.5663 864.1101,-385.1627"/>
+<polygon fill="#191970" stroke="#191970" points="864.4878,-381.6668 853.9859,-383.0671 863.0689,-388.5215 864.4878,-381.6668"/>
</g>
<!-- Node39 -->
<g id="node40" class="node">
<title>Node39</title>
<g id="a_node40"><a xlink:href="packed__func_8h.html" target="_top" xlink:title="Type-erased function used across TVM API. ">
-<polygon fill="#ffffff" stroke="#000000" points="1010,-425.5 1010,-455.5 1126,-455.5 1126,-425.5 1010,-425.5"/>
-<text text-anchor="start" x="1018" y="-443.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/packed</text>
-<text text-anchor="middle" x="1068" y="-432.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_func.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1698,-425.5 1698,-455.5 1814,-455.5 1814,-425.5 1698,-425.5"/>
+<text text-anchor="start" x="1706" y="-443.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/packed</text>
+<text text-anchor="middle" x="1756" y="-432.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_func.h</text>
</a>
</g>
</g>
<!-- Node4->Node39 -->
<g id="edge90" class="edge">
<title>Node4->Node39</title>
-<path fill="none" stroke="#191970" d="M696.5226,-493.3839C776.326,-482.023 916.3631,-462.0872 999.8188,-450.2064"/>
-<polygon fill="#191970" stroke="#191970" points="1000.5507,-453.6375 1009.9576,-448.763 999.5641,-446.7074 1000.5507,-453.6375"/>
+<path fill="none" stroke="#191970" d="M1870.75,-492.3906C1851.4919,-483.682 1822.5378,-470.5888 1798.5329,-459.7336"/>
+<polygon fill="#191970" stroke="#191970" points="1799.8232,-456.4759 1789.2693,-455.5446 1796.9389,-462.8541 1799.8232,-456.4759"/>
</g>
<!-- Node6 -->
<g id="node7" class="node">
<title>Node6</title>
<g id="a_node7"><a xlink:href="functor_8h.html" target="_top" xlink:title="Defines the Functor data structures. ">
-<polygon fill="#ffffff" stroke="#000000" points="2088.5,-297 2088.5,-316 2197.5,-316 2197.5,-297 2088.5,-297"/>
-<text text-anchor="middle" x="2143" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/functor.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2046.5,-297 2046.5,-316 2155.5,-316 2155.5,-297 2046.5,-297"/>
+<text text-anchor="middle" x="2101" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/functor.h</text>
</a>
</g>
</g>
<!-- Node5->Node6 -->
<g id="edge6" class="edge">
<title>Node5->Node6</title>
-<path fill="none" stroke="#191970" d="M735.6019,-367.8484C771.2606,-364.4875 818.2815,-360.4203 860,-358 1384.2512,-327.5856 1518.023,-369.5329 2041,-322 2053.1243,-320.898 2066.0007,-319.2886 2078.3752,-317.5188"/>
-<polygon fill="#191970" stroke="#191970" points="2079.2443,-320.9284 2088.6242,-316.0021 2078.2195,-314.0038 2079.2443,-320.9284"/>
+<path fill="none" stroke="#191970" d="M2172.1238,-358.2967C2157.2331,-347.4524 2137.2757,-332.9182 2122.2919,-322.006"/>
+<polygon fill="#191970" stroke="#191970" points="2124.2643,-319.1127 2114.1203,-316.055 2120.1434,-324.7712 2124.2643,-319.1127"/>
</g>
<!-- Node5->Node14 -->
<g id="edge43" class="edge">
<title>Node5->Node14</title>
-<path fill="none" stroke="#191970" d="M640.7233,-358.4356C615.8485,-348.4728 582.8069,-334.9127 554,-322 491.7846,-294.112 471.1265,-295.1194 416,-255 346.5418,-204.4504 280.9776,-124.7123 253.5265,-89.3757"/>
-<polygon fill="#191970" stroke="#191970" points="255.9732,-86.8155 247.1048,-81.0187 250.4227,-91.0807 255.9732,-86.8155"/>
+<path fill="none" stroke="#191970" d="M2249.5275,-367.1416C2292.1528,-360.5252 2350.9424,-347.406 2397,-322 2510.6513,-259.3083 2611.2358,-135.0353 2646.038,-89.1015"/>
+<polygon fill="#191970" stroke="#191970" points="2648.912,-91.1028 2652.1057,-81.0008 2643.3094,-86.9062 2648.912,-91.1028"/>
</g>
<!-- Node19 -->
<g id="node20" class="node">
<title>Node19</title>
<g id="a_node20"><a xlink:href="array_8h.html" target="_top" xlink:title="Runtime Array container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1314,-291.5 1314,-321.5 1440,-321.5 1440,-291.5 1314,-291.5"/>
-<text text-anchor="start" x="1322" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
-<text text-anchor="middle" x="1377" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/array.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1712,-291.5 1712,-321.5 1838,-321.5 1838,-291.5 1712,-291.5"/>
+<text text-anchor="start" x="1720" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
+<text text-anchor="middle" x="1775" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/array.h</text>
</a>
</g>
</g>
<!-- Node5->Node19 -->
<g id="edge21" class="edge">
<title>Node5->Node19</title>
-<path fill="none" stroke="#191970" d="M735.6536,-368.5796C771.3319,-365.4979 818.3554,-361.4676 860,-358 1054.4568,-341.8082 1106.9657,-350.6369 1303.7771,-322.0607"/>
-<polygon fill="#191970" stroke="#191970" points="1304.4829,-325.4947 1313.8694,-320.5811 1303.4675,-318.5687 1304.4829,-325.4947"/>
+<path fill="none" stroke="#191970" d="M2136.1916,-364.3944C2061.514,-352.4245 1930.2175,-331.3794 1848.1762,-318.2292"/>
+<polygon fill="#191970" stroke="#191970" points="1848.6032,-314.753 1838.1753,-316.6262 1847.4953,-321.6648 1848.6032,-314.753"/>
</g>
<!-- Node5->Node26 -->
<g id="edge38" class="edge">
<title>Node5->Node26</title>
-<path fill="none" stroke="#191970" d="M684.446,-358.2967C687.9312,-348.5672 692.4803,-335.8675 696.2018,-325.4784"/>
-<polygon fill="#191970" stroke="#191970" points="699.5,-326.6495 699.5773,-316.055 692.91,-324.2889 699.5,-326.6495"/>
+<path fill="none" stroke="#191970" d="M2221.2769,-358.4639C2242.418,-347.2222 2271.1458,-331.9463 2291.9784,-320.8686"/>
+<polygon fill="#191970" stroke="#191970" points="2293.7528,-323.8892 2300.9389,-316.1039 2290.4663,-317.7087 2293.7528,-323.8892"/>
</g>
<!-- Node7 -->
<g id="node8" class="node">
<title>Node7</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1710.5,-179.5 1710.5,-198.5 1799.5,-198.5 1799.5,-179.5 1710.5,-179.5"/>
-<text text-anchor="middle" x="1755" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/logging.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1388.5,-179.5 1388.5,-198.5 1477.5,-198.5 1477.5,-179.5 1388.5,-179.5"/>
+<text text-anchor="middle" x="1433" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/logging.h</text>
</g>
<!-- Node6->Node7 -->
<g id="edge7" class="edge">
<title>Node6->Node7</title>
-<path fill="none" stroke="#191970" d="M2130.0035,-296.9192C2108.2228,-281.3591 2062.4542,-250.7527 2019,-235 1949.3518,-209.7517 1864.2047,-198.1155 1809.7875,-192.9261"/>
-<polygon fill="#191970" stroke="#191970" points="1809.9383,-189.4254 1799.6614,-192.0012 1809.3015,-196.3964 1809.9383,-189.4254"/>
+<path fill="none" stroke="#191970" d="M2074.7634,-296.9592C2031.0513,-281.4568 1940.2813,-250.9313 1861,-235 1728.553,-208.3853 1569.7474,-196.421 1487.8483,-191.6875"/>
+<polygon fill="#191970" stroke="#191970" points="1487.9513,-188.1879 1477.77,-191.1193 1487.5571,-195.1768 1487.9513,-188.1879"/>
</g>
<!-- Node6->Node8 -->
<g id="edge8" class="edge">
<title>Node6->Node8</title>
-<path fill="none" stroke="#191970" d="M2143.5332,-296.5859C2143.8119,-281.6436 2142.2363,-253.1711 2128,-235 2089.234,-185.5195 2020.4614,-158.6683 1971.7697,-145.1206"/>
-<polygon fill="#191970" stroke="#191970" points="1972.5734,-141.7125 1962.0084,-142.5069 1970.7629,-148.4744 1972.5734,-141.7125"/>
+<path fill="none" stroke="#191970" d="M2092.2518,-296.9527C2068.9906,-272.2908 2002.5521,-206.4732 1932,-179 1859.7176,-150.853 1639.3493,-139.2881 1525.589,-135.0943"/>
+<polygon fill="#191970" stroke="#191970" points="1525.7065,-131.5964 1515.5871,-134.7345 1525.4548,-138.5918 1525.7065,-131.5964"/>
</g>
<!-- Node6->Node15 -->
<g id="edge18" class="edge">
<title>Node6->Node15</title>
-<path fill="none" stroke="#191970" d="M2134.1191,-296.9881C2119.1295,-281.5272 2087.1831,-251.0601 2054,-235 1953.4592,-186.34 1919.3289,-196.4296 1809,-179 1352.8434,-106.9371 1232.9847,-138.039 774,-87 750.4857,-84.3852 724.3625,-80.9422 702.8688,-77.9654"/>
-<polygon fill="#191970" stroke="#191970" points="703.1093,-74.4651 692.721,-76.5472 702.1403,-81.3977 703.1093,-74.4651"/>
+<path fill="none" stroke="#191970" d="M2108.9666,-296.9967C2138.8513,-261.3474 2244.5596,-135.2487 2283.4779,-88.8233"/>
+<polygon fill="#191970" stroke="#191970" points="2286.1922,-91.0336 2289.9342,-81.1216 2280.8277,-86.5366 2286.1922,-91.0336"/>
</g>
<!-- Node6->Node16 -->
<g id="edge19" class="edge">
<title>Node6->Node16</title>
-<path fill="none" stroke="#191970" d="M2148.7291,-296.9655C2158.7687,-281.1856 2181.1466,-249.9042 2209,-235 2299.0661,-186.806 2369.6669,-276.6819 2436,-199 2457.9343,-173.313 2445.4947,-155.4159 2436,-123 2432.426,-110.798 2424.9763,-98.7442 2418.0582,-89.3692"/>
-<polygon fill="#191970" stroke="#191970" points="2420.6509,-87.003 2411.732,-81.2842 2415.1379,-91.3167 2420.6509,-87.003"/>
+<path fill="none" stroke="#191970" d="M2095.8223,-296.8627C2087.666,-282.282 2070.7092,-254.2754 2051,-235 2018.9217,-203.6278 2009.042,-194.6748 1967,-179 1865.9633,-141.3297 1085.187,-85.5513 914.778,-73.7452"/>
+<polygon fill="#191970" stroke="#191970" points="914.8864,-70.2444 904.669,-73.0469 914.404,-77.2278 914.8864,-70.2444"/>
</g>
<!-- Node6->Node18 -->
<g id="edge20" class="edge">
<title>Node6->Node18</title>
-<path fill="none" stroke="#191970" d="M2118.4688,-296.8906C2091.7369,-286.4192 2048.8218,-269.6085 2019.1151,-257.9718"/>
-<polygon fill="#191970" stroke="#191970" points="2020.1771,-254.629 2009.5894,-254.2404 2017.624,-261.1467 2020.1771,-254.629"/>
+<path fill="none" stroke="#191970" d="M2046.3087,-298.5859C1957.2034,-285.6918 1783.2892,-260.5254 1709.9315,-249.9101"/>
+<polygon fill="#191970" stroke="#191970" points="1710.2581,-246.421 1699.8599,-248.4527 1709.2555,-253.3488 1710.2581,-246.421"/>
</g>
<!-- Node8->Node9 -->
<g id="edge9" class="edge">
<title>Node8->Node9</title>
-<path fill="none" stroke="#191970" d="M1856.4531,-127.2901C1839.5134,-125.7725 1821.038,-124.2187 1804,-123 1475.7558,-99.5215 1083.481,-81.5815 922.0404,-74.6361"/>
-<polygon fill="#191970" stroke="#191970" points="921.8843,-71.1263 911.7435,-74.1946 921.5844,-78.1199 921.8843,-71.1263"/>
+<path fill="none" stroke="#191970" d="M1396.2137,-130.0957C1188.0461,-119.9833 495.2122,-86.3267 265.0543,-75.146"/>
+<polygon fill="#191970" stroke="#191970" points="264.9315,-71.636 254.7734,-74.6466 264.5918,-78.6278 264.9315,-71.636"/>
</g>
<!-- Node13 -->
<g id="node14" class="node">
<title>Node13</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1765.5,-62 1765.5,-81 1890.5,-81 1890.5,-62 1765.5,-62"/>
-<text text-anchor="middle" x="1828" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/logging.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1481.5,-62 1481.5,-81 1606.5,-81 1606.5,-62 1481.5,-62"/>
+<text text-anchor="middle" x="1544" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/logging.h</text>
</g>
<!-- Node8->Node13 -->
<g id="edge13" class="edge">
<title>Node8->Node13</title>
-<path fill="none" stroke="#191970" d="M1902.25,-123.3906C1888.3009,-113.6421 1866.4902,-98.3994 1850.1103,-86.9521"/>
-<polygon fill="#191970" stroke="#191970" points="1851.8551,-83.9015 1841.6534,-81.0419 1847.8452,-89.6392 1851.8551,-83.9015"/>
+<path fill="none" stroke="#191970" d="M1469.75,-123.3906C1483.6991,-113.6421 1505.5098,-98.3994 1521.8897,-86.9521"/>
+<polygon fill="#191970" stroke="#191970" points="1524.1548,-89.6392 1530.3466,-81.0419 1520.1449,-83.9015 1524.1548,-89.6392"/>
</g>
<!-- Node8->Node14 -->
<g id="edge14" class="edge">
<title>Node8->Node14</title>
-<path fill="none" stroke="#191970" d="M1856.3443,-131.0175C1671.7974,-124.8573 1093.7876,-105.3591 615,-87 488.2512,-82.1398 337.433,-75.7087 272.4381,-72.906"/>
-<polygon fill="#191970" stroke="#191970" points="272.5419,-69.4073 262.4001,-72.4726 272.2398,-76.4008 272.5419,-69.4073"/>
+<path fill="none" stroke="#191970" d="M1515.657,-129.9502C1730.9312,-118.9449 2463.6152,-81.4885 2626.6663,-73.153"/>
+<polygon fill="#191970" stroke="#191970" points="2627.182,-76.6313 2636.9902,-72.6252 2626.8245,-69.6404 2627.182,-76.6313"/>
</g>
<!-- Node8->Node15 -->
<g id="edge15" class="edge">
<title>Node8->Node15</title>
-<path fill="none" stroke="#191970" d="M1856.467,-127.0816C1839.5275,-125.5614 1821.0488,-124.0583 1804,-123 1346.8226,-94.6217 1230.4466,-125.3793 774,-87 750.424,-85.0177 724.2951,-81.6317 702.8138,-78.5293"/>
-<polygon fill="#191970" stroke="#191970" points="703.0764,-75.0303 692.6733,-77.0364 702.0568,-81.9556 703.0764,-75.0303"/>
+<path fill="none" stroke="#191970" d="M1515.5433,-128.6509C1675.4261,-116.973 2111.5654,-85.1173 2253.0407,-74.7838"/>
+<polygon fill="#191970" stroke="#191970" points="2253.381,-78.2684 2263.0995,-74.0491 2252.871,-71.287 2253.381,-78.2684"/>
</g>
<!-- Node8->Node16 -->
<g id="edge16" class="edge">
<title>Node8->Node16</title>
-<path fill="none" stroke="#191970" d="M1975.7158,-125.4589C2079.3879,-112.3668 2289.3231,-85.8555 2370.2035,-75.6417"/>
-<polygon fill="#191970" stroke="#191970" points="2370.8132,-79.0925 2380.2959,-74.3672 2369.9361,-72.1477 2370.8132,-79.0925"/>
+<path fill="none" stroke="#191970" d="M1396.4653,-126.6213C1276.2034,-113.7361 1007.8344,-84.9823 914.6353,-74.9966"/>
+<polygon fill="#191970" stroke="#191970" points="914.9747,-71.5131 904.6587,-73.9277 914.2289,-78.4732 914.9747,-71.5131"/>
</g>
<!-- Node17 -->
<g id="node18" class="node">
<title>Node17</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2120,-62 2120,-81 2170,-81 2170,-62 2120,-62"/>
-<text text-anchor="middle" x="2145" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">atomic</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1187,-62 1187,-81 1237,-81 1237,-62 1187,-62"/>
+<text text-anchor="middle" x="1212" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">atomic</text>
</g>
<!-- Node8->Node17 -->
<g id="edge17" class="edge">
<title>Node8->Node17</title>
-<path fill="none" stroke="#191970" d="M1951.7813,-123.3906C1994.9869,-111.7874 2067.1761,-92.4003 2110.0198,-80.8943"/>
-<polygon fill="#191970" stroke="#191970" points="2110.9781,-84.261 2119.7281,-78.287 2109.1625,-77.5005 2110.9781,-84.261"/>
+<path fill="none" stroke="#191970" d="M1418.1427,-123.4581C1371.3511,-111.6643 1292.3626,-91.7553 1246.9524,-80.3097"/>
+<polygon fill="#191970" stroke="#191970" points="1247.8061,-76.9155 1237.2539,-77.8652 1246.0952,-83.7032 1247.8061,-76.9155"/>
</g>
<!-- Node10 -->
<g id="node11" class="node">
<title>Node10</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="707.5,-.5 707.5,-19.5 800.5,-19.5 800.5,-.5 707.5,-.5"/>
-<text text-anchor="middle" x="754" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dlpack/dlpack.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="50.5,-.5 50.5,-19.5 143.5,-19.5 143.5,-.5 50.5,-.5"/>
+<text text-anchor="middle" x="97" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dlpack/dlpack.h</text>
</g>
<!-- Node9->Node10 -->
<g id="edge10" class="edge">
<title>Node9->Node10</title>
-<path fill="none" stroke="#191970" d="M824.0112,-56.2977C809.8034,-46.9022 791.6215,-34.8787 777.3475,-25.4395"/>
-<polygon fill="#191970" stroke="#191970" points="778.8845,-22.2598 768.6128,-19.6633 775.0233,-28.0986 778.8845,-22.2598"/>
+<path fill="none" stroke="#191970" d="M167.0112,-56.2977C152.8034,-46.9022 134.6215,-34.8787 120.3475,-25.4395"/>
+<polygon fill="#191970" stroke="#191970" points="121.8845,-22.2598 111.6128,-19.6633 118.0233,-28.0986 121.8845,-22.2598"/>
</g>
<!-- Node11 -->
<g id="node12" class="node">
<title>Node11</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="819,-.5 819,-19.5 875,-19.5 875,-.5 819,-.5"/>
-<text text-anchor="middle" x="847" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stddef.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="162,-.5 162,-19.5 218,-19.5 218,-.5 162,-.5"/>
+<text text-anchor="middle" x="190" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stddef.h</text>
</g>
<!-- Node9->Node11 -->
<g id="edge11" class="edge">
<title>Node9->Node11</title>
-<path fill="none" stroke="#191970" d="M847,-56.2977C847,-48.3834 847,-38.6043 847,-30.0759"/>
-<polygon fill="#191970" stroke="#191970" points="850.5001,-29.8469 847,-19.8469 843.5001,-29.847 850.5001,-29.8469"/>
+<path fill="none" stroke="#191970" d="M190,-56.2977C190,-48.3834 190,-38.6043 190,-30.0759"/>
+<polygon fill="#191970" stroke="#191970" points="193.5001,-29.8469 190,-19.8469 186.5001,-29.847 193.5001,-29.8469"/>
</g>
<!-- Node12 -->
<g id="node13" class="node">
<title>Node12</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="893.5,-.5 893.5,-19.5 946.5,-19.5 946.5,-.5 893.5,-.5"/>
-<text text-anchor="middle" x="920" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stdint.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="236.5,-.5 236.5,-19.5 289.5,-19.5 289.5,-.5 236.5,-.5"/>
+<text text-anchor="middle" x="263" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stdint.h</text>
</g>
<!-- Node9->Node12 -->
<g id="edge12" class="edge">
<title>Node9->Node12</title>
-<path fill="none" stroke="#191970" d="M865.0449,-56.2977C875.8114,-47.2274 889.4851,-35.7077 900.4995,-26.4285"/>
-<polygon fill="#191970" stroke="#191970" points="902.9191,-28.9667 908.3118,-19.8469 898.4089,-23.6132 902.9191,-28.9667"/>
+<path fill="none" stroke="#191970" d="M208.0449,-56.2977C218.8114,-47.2274 232.4851,-35.7077 243.4995,-26.4285"/>
+<polygon fill="#191970" stroke="#191970" points="245.9191,-28.9667 251.3118,-19.8469 241.4089,-23.6132 245.9191,-28.9667"/>
</g>
<!-- Node19->Node16 -->
<g id="edge24" class="edge">
<title>Node19->Node16</title>
-<path fill="none" stroke="#191970" d="M1440.2629,-302.2871C1528.8635,-295.7265 1694.6919,-281.0371 1834,-255 1868.3333,-248.583 1875.9547,-242.8017 1910,-235 1991.223,-216.3872 2014.2056,-223.0095 2094,-199 2200.4628,-166.9661 2320.6299,-111.4745 2374.369,-85.5737"/>
-<polygon fill="#191970" stroke="#191970" points="2376.1395,-88.6051 2383.6119,-81.0941 2373.0866,-82.3059 2376.1395,-88.6051"/>
+<path fill="none" stroke="#191970" d="M1769.3115,-291.1724C1757.9141,-262.9702 1729.3282,-203.5028 1683,-179 1565.8511,-117.0403 1518.3922,-160.2899 1387,-143 1208.4598,-119.5059 994.8885,-88.2169 914.6189,-76.3446"/>
+<polygon fill="#191970" stroke="#191970" points="915.0332,-72.8678 904.6284,-74.8653 914.0079,-79.7924 915.0332,-72.8678"/>
</g>
<!-- Node19->Node18 -->
<g id="edge25" class="edge">
<title>Node19->Node18</title>
-<path fill="none" stroke="#191970" d="M1440.251,-301.0999C1540.1439,-292.4112 1740.5071,-274.3335 1910,-255 1923.8763,-253.4172 1939.1477,-251.4516 1952.3383,-249.683"/>
-<polygon fill="#191970" stroke="#191970" points="1952.9656,-253.1301 1962.4045,-248.318 1952.0249,-246.1935 1952.9656,-253.1301"/>
+<path fill="none" stroke="#191970" d="M1750.7819,-291.4554C1735.4227,-281.9141 1715.6243,-269.6151 1700.2575,-260.0691"/>
+<polygon fill="#191970" stroke="#191970" points="1701.6828,-256.8341 1691.3415,-254.5303 1697.989,-262.7802 1701.6828,-256.8341"/>
</g>
<!-- Node19->Node20 -->
<g id="edge22" class="edge">
<title>Node19->Node20</title>
-<path fill="none" stroke="#191970" d="M1440.1505,-292.3175C1487.1965,-281.2544 1547.0906,-265.975 1570,-255 1597.3311,-241.9068 1624.8604,-220.1763 1642.1437,-205.2501"/>
-<polygon fill="#191970" stroke="#191970" points="1644.5722,-207.7751 1649.7594,-198.5369 1639.9433,-202.524 1644.5722,-207.7751"/>
+<path fill="none" stroke="#191970" d="M1762.4842,-291.3903C1749.6091,-276.2868 1728.583,-252.8095 1708,-235 1694.9964,-223.7486 1679.2352,-212.6801 1666.3764,-204.2208"/>
+<polygon fill="#191970" stroke="#191970" points="1667.9063,-201.0424 1657.6071,-198.557 1664.1085,-206.9226 1667.9063,-201.0424"/>
</g>
<!-- Node21 -->
<g id="node22" class="node">
<title>Node21</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1026,-235.5 1026,-254.5 1084,-254.5 1084,-235.5 1026,-235.5"/>
-<text text-anchor="middle" x="1055" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">memory</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1794,-235.5 1794,-254.5 1852,-254.5 1852,-235.5 1794,-235.5"/>
+<text text-anchor="middle" x="1823" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">memory</text>
</g>
<!-- Node19->Node21 -->
<g id="edge23" class="edge">
<title>Node19->Node21</title>
-<path fill="none" stroke="#191970" d="M1313.8062,-292.7834C1310.8309,-292.1729 1307.884,-291.5756 1305,-291 1230.6569,-276.1617 1143.5182,-260.5247 1094.2304,-251.842"/>
-<polygon fill="#191970" stroke="#191970" points="1094.7606,-248.3817 1084.3057,-250.0976 1093.5488,-255.276 1094.7606,-248.3817"/>
+<path fill="none" stroke="#191970" d="M1786.8652,-291.2977C1793.5975,-282.672 1802.0585,-271.8312 1809.1006,-262.8087"/>
+<polygon fill="#191970" stroke="#191970" points="1811.921,-264.8836 1815.3146,-254.8469 1806.4027,-260.5766 1811.921,-264.8836"/>
</g>
<!-- Node22 -->
<g id="node23" class="node">
<title>Node22</title>
<g id="a_node23"><a xlink:href="runtime_2container_2base_8h.html" target="_top" xlink:title="Base utilities for common POD(plain old data) container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1736.5,-235.5 1736.5,-254.5 1791.5,-254.5 1791.5,-235.5 1736.5,-235.5"/>
-<text text-anchor="middle" x="1764" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">./base.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1312.5,-235.5 1312.5,-254.5 1367.5,-254.5 1367.5,-235.5 1312.5,-235.5"/>
+<text text-anchor="middle" x="1340" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">./base.h</text>
</a>
</g>
</g>
<!-- Node19->Node22 -->
<g id="edge26" class="edge">
<title>Node19->Node22</title>
-<path fill="none" stroke="#191970" d="M1440.0395,-296.4821C1521.1584,-283.5911 1659.5343,-261.6011 1726.0711,-251.0275"/>
-<polygon fill="#191970" stroke="#191970" points="1726.7865,-254.4578 1736.1132,-249.4316 1725.6878,-247.5445 1726.7865,-254.4578"/>
+<path fill="none" stroke="#191970" d="M1711.7473,-297.5574C1619.9261,-284.5758 1452.5403,-260.9109 1377.7975,-250.3438"/>
+<polygon fill="#191970" stroke="#191970" points="1378.2749,-246.8765 1367.8833,-248.9421 1377.2949,-253.8076 1378.2749,-246.8765"/>
</g>
<!-- Node22->Node7 -->
<g id="edge27" class="edge">
<title>Node22->Node7</title>
-<path fill="none" stroke="#191970" d="M1762.4323,-235.2455C1761.2579,-227.9382 1759.6116,-217.6944 1758.1668,-208.7046"/>
-<polygon fill="#191970" stroke="#191970" points="1761.5922,-207.9606 1756.5497,-198.6427 1754.6809,-209.0714 1761.5922,-207.9606"/>
+<path fill="none" stroke="#191970" d="M1356.1994,-235.2455C1370.6018,-226.5731 1391.8732,-213.7645 1408.3799,-203.825"/>
+<polygon fill="#191970" stroke="#191970" points="1410.225,-206.7996 1416.9863,-198.6427 1406.614,-200.8028 1410.225,-206.7996"/>
</g>
<!-- Node22->Node8 -->
<g id="edge34" class="edge">
<title>Node22->Node8</title>
-<path fill="none" stroke="#191970" d="M1782.6037,-235.3689C1798.8202,-226.6645 1822.6994,-213.0895 1842,-199 1862.8821,-183.756 1884.7318,-163.6824 1899.2266,-149.6927"/>
-<polygon fill="#191970" stroke="#191970" points="1901.763,-152.1078 1906.4697,-142.6158 1896.871,-147.1009 1901.763,-152.1078"/>
+<path fill="none" stroke="#191970" d="M1343.8027,-235.4928C1349.685,-221.7116 1362.1007,-195.9806 1379,-179 1392.099,-165.8379 1409.7026,-154.9567 1424.7717,-147.0874"/>
+<polygon fill="#191970" stroke="#191970" points="1426.3903,-150.1913 1433.764,-142.5834 1423.2554,-143.9325 1426.3903,-150.1913"/>
</g>
<!-- Node22->Node13 -->
<g id="edge28" class="edge">
<title>Node22->Node13</title>
-<path fill="none" stroke="#191970" d="M1777.3409,-235.2528C1787.7516,-226.8681 1801.6422,-213.8072 1809,-199 1826.4371,-163.9087 1828.91,-117.1764 1828.7313,-91.1418"/>
-<polygon fill="#191970" stroke="#191970" points="1832.2286,-90.972 1828.5177,-81.0483 1825.2302,-91.1202 1832.2286,-90.972"/>
+<path fill="none" stroke="#191970" d="M1340.7522,-235.3811C1343.0935,-212.283 1352.3364,-153.3564 1387,-123 1410.4548,-102.4596 1442.2925,-90.0776 1471.1675,-82.6289"/>
+<polygon fill="#191970" stroke="#191970" points="1472.3409,-85.9461 1481.2387,-80.1946 1470.6963,-79.142 1472.3409,-85.9461"/>
</g>
<!-- Node22->Node16 -->
<g id="edge37" class="edge">
<title>Node22->Node16</title>
-<path fill="none" stroke="#191970" d="M1785.8887,-235.3824C1845.3654,-209.3743 2009.287,-138.5607 2066,-123 2175.0338,-93.0837 2308.9955,-79.1383 2369.9685,-73.9869"/>
-<polygon fill="#191970" stroke="#191970" points="2370.4603,-77.4584 2380.1406,-73.1525 2369.888,-70.4818 2370.4603,-77.4584"/>
+<path fill="none" stroke="#191970" d="M1331.637,-235.4291C1318.5957,-221.0883 1291.9919,-194.1002 1264,-179 1144.8584,-114.7294 983.5684,-85.7894 915.0175,-75.8075"/>
+<polygon fill="#191970" stroke="#191970" points="915.2244,-72.302 904.8325,-74.3657 914.2432,-79.2329 915.2244,-72.302"/>
</g>
<!-- Node22->Node20 -->
<g id="edge35" class="edge">
<title>Node22->Node20</title>
-<path fill="none" stroke="#191970" d="M1745.8846,-235.2455C1729.555,-226.4527 1705.329,-213.4079 1686.7661,-203.4125"/>
-<polygon fill="#191970" stroke="#191970" points="1688.3719,-200.3021 1677.9078,-198.6427 1685.0532,-206.4654 1688.3719,-200.3021"/>
+<path fill="none" stroke="#191970" d="M1367.6011,-239.8819C1420.5666,-230.0605 1536.7731,-208.5123 1599.6278,-196.8571"/>
+<polygon fill="#191970" stroke="#191970" points="1600.4148,-200.2709 1609.6091,-195.0063 1599.1385,-193.3882 1600.4148,-200.2709"/>
</g>
<!-- Node22->Node23 -->
<g id="edge29" class="edge">
<title>Node22->Node23</title>
-<path fill="none" stroke="#191970" d="M1736.4648,-243.3544C1611.5873,-235.891 1096.1539,-205.086 901.7031,-193.4646"/>
-<polygon fill="#191970" stroke="#191970" points="901.7687,-189.9624 891.5777,-192.8595 901.3511,-196.9499 901.7687,-189.9624"/>
+<path fill="none" stroke="#191970" d="M1312.049,-238.0417C1307.3709,-236.9643 1302.5574,-235.9091 1298,-235 1227.058,-220.8494 1145.5163,-208.0329 1087.878,-199.5442"/>
+<polygon fill="#191970" stroke="#191970" points="1088.1635,-196.0488 1077.7619,-198.0627 1087.1491,-202.9749 1088.1635,-196.0488"/>
</g>
<!-- Node25 -->
<g id="node26" class="node">
<title>Node25</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1517.5,-179.5 1517.5,-198.5 1600.5,-198.5 1600.5,-179.5 1517.5,-179.5"/>
-<text text-anchor="middle" x="1559" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">initializer_list</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1172.5,-179.5 1172.5,-198.5 1255.5,-198.5 1255.5,-179.5 1172.5,-179.5"/>
+<text text-anchor="middle" x="1214" y="-186.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">initializer_list</text>
</g>
<!-- Node22->Node25 -->
<g id="edge36" class="edge">
<title>Node22->Node25</title>
-<path fill="none" stroke="#191970" d="M1736.3528,-237.4476C1702.3996,-228.1726 1644.3524,-212.3158 1603.9518,-201.2795"/>
-<polygon fill="#191970" stroke="#191970" points="1604.6549,-197.8434 1594.086,-198.5845 1602.8102,-204.596 1604.6549,-197.8434"/>
+<path fill="none" stroke="#191970" d="M1318.3398,-235.3733C1298.0228,-226.3435 1267.436,-212.7494 1244.625,-202.6111"/>
+<polygon fill="#191970" stroke="#191970" points="1245.9372,-199.3642 1235.3775,-198.5011 1243.0941,-205.7609 1245.9372,-199.3642"/>
</g>
<!-- Node23->Node8 -->
<g id="edge30" class="edge">
<title>Node23->Node8</title>
-<path fill="none" stroke="#191970" d="M891.8606,-185.6646C1083.8148,-175.7937 1648.0057,-146.7812 1846.1958,-136.5896"/>
-<polygon fill="#191970" stroke="#191970" points="1846.3907,-140.0843 1856.1977,-136.0752 1846.0311,-133.0935 1846.3907,-140.0843"/>
+<path fill="none" stroke="#191970" d="M1077.6692,-180.8129C1082.5187,-180.1993 1087.3367,-179.5898 1092,-179 1194.4524,-166.042 1313.1273,-151.0475 1386.3871,-141.7929"/>
+<polygon fill="#191970" stroke="#191970" points="1386.9341,-145.2517 1396.4166,-140.526 1386.0567,-138.3069 1386.9341,-145.2517"/>
</g>
<!-- Node23->Node15 -->
<g id="edge32" class="edge">
<title>Node23->Node15</title>
-<path fill="none" stroke="#191970" d="M813.17,-179.3845C783.9553,-159.0724 715.7377,-111.6431 680.2849,-86.9939"/>
-<polygon fill="#191970" stroke="#191970" points="682.2296,-84.0832 672.0211,-81.2484 678.2336,-89.8306 682.2296,-84.0832"/>
+<path fill="none" stroke="#191970" d="M1057.0817,-179.4748C1126.7681,-164.813 1266.8185,-136.9517 1387,-123 1717.9259,-84.5831 2119.9423,-74.5298 2252.9163,-72.1648"/>
+<polygon fill="#191970" stroke="#191970" points="2253.1826,-75.6608 2263.1209,-71.9891 2253.0621,-68.6618 2253.1826,-75.6608"/>
</g>
<!-- Node23->Node16 -->
<g id="edge33" class="edge">
<title>Node23->Node16</title>
-<path fill="none" stroke="#191970" d="M891.7192,-184.1748C1159.3292,-164.2229 2174.5741,-88.5305 2370.2526,-73.9415"/>
-<polygon fill="#191970" stroke="#191970" points="2370.7493,-77.4143 2380.4614,-73.1804 2370.2288,-70.4337 2370.7493,-77.4143"/>
+<path fill="none" stroke="#191970" d="M1002.2797,-179.3845C979.9218,-159.3306 928.0947,-112.8445 900.3343,-87.9449"/>
+<polygon fill="#191970" stroke="#191970" points="902.6497,-85.32 892.8684,-81.2484 897.9757,-90.531 902.6497,-85.32"/>
</g>
<!-- Node24 -->
<g id="node25" class="node">
<title>Node24</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="802.5,-123.5 802.5,-142.5 851.5,-142.5 851.5,-123.5 802.5,-123.5"/>
-<text text-anchor="middle" x="827" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstdlib</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="988.5,-123.5 988.5,-142.5 1037.5,-142.5 1037.5,-123.5 988.5,-123.5"/>
+<text text-anchor="middle" x="1013" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstdlib</text>
</g>
<!-- Node23->Node24 -->
<g id="edge31" class="edge">
<title>Node23->Node24</title>
-<path fill="none" stroke="#191970" d="M827,-179.2455C827,-171.9382 827,-161.6944 827,-152.7046"/>
-<polygon fill="#191970" stroke="#191970" points="830.5001,-152.6426 827,-142.6427 823.5001,-152.6427 830.5001,-152.6426"/>
+<path fill="none" stroke="#191970" d="M1013,-179.2455C1013,-171.9382 1013,-161.6944 1013,-152.7046"/>
+<polygon fill="#191970" stroke="#191970" points="1016.5001,-152.6426 1013,-142.6427 1009.5001,-152.6427 1016.5001,-152.6426"/>
</g>
<!-- Node26->Node9 -->
<g id="edge39" class="edge">
<title>Node26->Node9</title>
-<path fill="none" stroke="#191970" d="M702.7997,-296.8566C702.6607,-275.0769 704.0838,-220.7304 720,-179 730.7188,-150.8966 735.9837,-143.4933 758,-123 771.4355,-110.4939 788.4357,-99.726 803.9696,-91.3174"/>
-<polygon fill="#191970" stroke="#191970" points="805.6682,-94.3789 812.9083,-86.6439 802.4248,-88.1756 805.6682,-94.3789"/>
+<path fill="none" stroke="#191970" d="M2249.6304,-298.698C2222.8968,-295.9184 2192.0554,-292.9876 2164,-291 1781.9566,-263.9337 1681.7257,-312.0667 1303,-255 1268.4621,-249.7958 1261.2937,-241.6252 1227,-235 1100.3458,-210.5316 1066.3712,-219.4102 939,-199 765.4921,-171.1968 724.2602,-152.3072 551,-123 452.3333,-106.3104 337.8825,-90.5969 264.7546,-81.0168"/>
+<polygon fill="#191970" stroke="#191970" points="265.0851,-77.5303 254.7161,-79.7063 264.1788,-84.4714 265.0851,-77.5303"/>
</g>
<!-- Node26->Node13 -->
<g id="edge40" class="edge">
<title>Node26->Node13</title>
-<path fill="none" stroke="#191970" d="M772.2227,-300.5582C850.6108,-293.0913 981.8844,-278.3491 1093,-255 1298.6423,-211.7877 1342.5015,-171.3363 1547,-123 1617.4629,-106.345 1698.9405,-91.9379 1755.9567,-82.631"/>
-<polygon fill="#191970" stroke="#191970" points="1756.6434,-86.0655 1765.9544,-81.0104 1755.5232,-79.1557 1756.6434,-86.0655"/>
+<path fill="none" stroke="#191970" d="M2308.6467,-296.6593C2274.48,-264.866 2160.9255,-164.5627 2046,-123 1968.9473,-95.134 1736.4949,-80.4952 1616.8662,-74.631"/>
+<polygon fill="#191970" stroke="#191970" points="1616.8574,-71.1266 1606.7003,-74.1405 1616.52,-78.1185 1616.8574,-71.1266"/>
</g>
<!-- Node26->Node14 -->
<g id="edge41" class="edge">
<title>Node26->Node14</title>
-<path fill="none" stroke="#191970" d="M689.7709,-296.8902C653.8186,-270.8376 554.6933,-199.4471 520,-179 471.3681,-150.338 458.4497,-143.871 406,-123 360.3927,-104.8518 305.719,-88.973 271.9165,-79.8073"/>
-<polygon fill="#191970" stroke="#191970" points="272.6946,-76.3924 262.1293,-77.1833 270.8818,-83.1536 272.6946,-76.3924"/>
+<path fill="none" stroke="#191970" d="M2332.7495,-296.9967C2385.0762,-260.8297 2572.0946,-131.567 2636.7893,-86.8515"/>
+<polygon fill="#191970" stroke="#191970" points="2638.8432,-89.6866 2645.0794,-81.1216 2634.8631,-83.9282 2638.8432,-89.6866"/>
</g>
<!-- Node26->Node15 -->
<g id="edge42" class="edge">
<title>Node26->Node15</title>
-<path fill="none" stroke="#191970" d="M685.3709,-296.8731C671.2442,-288.0568 652.4385,-273.5669 644,-255 618.9495,-199.8824 638.6546,-125.8214 650.5839,-91.0599"/>
-<polygon fill="#191970" stroke="#191970" points="654.0288,-91.8229 654.1173,-81.2285 647.4413,-89.4554 654.0288,-91.8229"/>
+<path fill="none" stroke="#191970" d="M2319.3975,-296.8759C2320.3648,-269.491 2322.0744,-188.8649 2312,-123 2310.3511,-112.2197 2307.2589,-100.4656 2304.3992,-90.912"/>
+<polygon fill="#191970" stroke="#191970" points="2307.6941,-89.722 2301.3586,-81.2301 2301.0156,-91.8194 2307.6941,-89.722"/>
</g>
<!-- Node27->Node6 -->
<g id="edge45" class="edge">
<title>Node27->Node6</title>
-<path fill="none" stroke="#191970" d="M1333.6286,-438.8691C1476.0252,-434.351 1846.1048,-419.912 1965,-389 2022.9059,-373.9448 2085.1531,-340.7068 2118.5199,-321.3276"/>
-<polygon fill="#191970" stroke="#191970" points="2120.6841,-324.115 2127.5249,-316.0246 2117.1319,-318.0832 2120.6841,-324.115"/>
+<path fill="none" stroke="#191970" d="M2006.0663,-425.389C2026.0024,-400.4222 2065.7169,-350.6864 2086.9094,-324.1462"/>
+<polygon fill="#191970" stroke="#191970" points="2089.8262,-326.1025 2093.331,-316.1042 2084.3561,-321.7346 2089.8262,-326.1025"/>
</g>
<!-- Node27->Node14 -->
<g id="edge84" class="edge">
<title>Node27->Node14</title>
-<path fill="none" stroke="#191970" d="M1220.2898,-433.3601C1133.5976,-422.1683 975.6296,-400.5797 952,-389 933.93,-380.1448 937.2235,-366.5348 919,-358 799.7842,-302.1665 747.3084,-370.6871 625,-322 603.4444,-313.4194 602.1148,-304.149 583,-291 466.0272,-210.535 321.5026,-121.2388 264.6339,-86.4824"/>
-<polygon fill="#191970" stroke="#191970" points="266.3303,-83.4175 255.971,-81.1958 262.6838,-89.3927 266.3303,-83.4175"/>
+<path fill="none" stroke="#191970" d="M2050.5333,-436.758C2124.4537,-430.9506 2257.5335,-417.3455 2368,-389 2405.3368,-379.4194 2412.3494,-369.9398 2449,-358 2507.2264,-339.0314 2533.3522,-359.1957 2582,-322 2632.3205,-283.5254 2636.2273,-259.7995 2654,-199 2664.7798,-162.1226 2663.2482,-116.6918 2661.139,-91.2237"/>
+<polygon fill="#191970" stroke="#191970" points="2664.6007,-90.6538 2660.1714,-81.0293 2657.632,-91.3153 2664.6007,-90.6538"/>
</g>
<!-- Node27->Node26 -->
<g id="edge46" class="edge">
<title>Node27->Node26</title>
-<path fill="none" stroke="#191970" d="M1220.3908,-433.1725C1150.1309,-423.559 1034.7213,-405.9174 995,-389 973.3243,-379.7683 973.5666,-367.4838 952,-358 943.2618,-354.1574 832.6302,-331.9912 761.8044,-318.0204"/>
-<polygon fill="#191970" stroke="#191970" points="762.1484,-314.521 751.6604,-316.0218 760.7952,-321.3889 762.1484,-314.521"/>
+<path fill="none" stroke="#191970" d="M2022.0203,-425.4588C2040.0112,-415.56 2063.7193,-402.062 2084,-389 2103.8071,-376.243 2105.718,-368.1063 2127,-358 2151.5208,-346.3556 2218.4057,-329.5755 2266.3305,-318.3683"/>
+<polygon fill="#191970" stroke="#191970" points="2267.2983,-321.7368 2276.2473,-316.0653 2265.7147,-314.9183 2267.2983,-321.7368"/>
</g>
<!-- Node27->Node28 -->
<g id="edge47" class="edge">
<title>Node27->Node28</title>
-<path fill="none" stroke="#191970" d="M1333.8825,-437.0067C1432.4405,-430.5212 1641.1799,-414.8765 1816,-389 1823.6825,-387.8628 1831.7578,-386.4645 1839.6709,-384.9775"/>
-<polygon fill="#191970" stroke="#191970" points="1840.4143,-388.3986 1849.5655,-383.0594 1839.0822,-381.5265 1840.4143,-388.3986"/>
+<path fill="none" stroke="#191970" d="M1937.4465,-434.6292C1904.1608,-431.36 1861.2255,-427.4643 1823,-425 1406.6684,-398.1601 1298.9302,-441.0971 885,-389 877.2396,-388.0233 869.0768,-386.6283 861.1471,-385.0707"/>
+<polygon fill="#191970" stroke="#191970" points="861.7768,-381.627 851.2752,-383.0301 860.3597,-388.4821 861.7768,-381.627"/>
</g>
<!-- Node38 -->
<g id="node39" class="node">
<title>Node38</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2558.5,-297 2558.5,-316 2623.5,-316 2623.5,-297 2558.5,-297"/>
-<text text-anchor="middle" x="2591" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">functional</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="856.5,-297 856.5,-316 921.5,-316 921.5,-297 856.5,-297"/>
+<text text-anchor="middle" x="889" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">functional</text>
</g>
<!-- Node27->Node38 -->
<g id="edge83" class="edge">
<title>Node27->Node38</title>
-<path fill="none" stroke="#191970" d="M1333.8773,-437.6728C1458.2961,-431.2426 1763.911,-414.1085 2019,-389 2253.0945,-365.958 2314.5855,-373.963 2544,-322 2547.6504,-321.1732 2551.4194,-320.1789 2555.161,-319.0981"/>
-<polygon fill="#191970" stroke="#191970" points="2556.3783,-322.3854 2564.9006,-316.0908 2554.3131,-315.697 2556.3783,-322.3854"/>
+<path fill="none" stroke="#191970" d="M1958.6576,-425.3988C1911.2577,-405.9286 1824.0123,-372.817 1746,-358 1456.0822,-302.9356 1377.4389,-341.753 1083,-322 1030.8549,-318.5017 970.8995,-313.5723 931.6084,-310.2182"/>
+<polygon fill="#191970" stroke="#191970" points="931.853,-306.7264 921.5905,-309.3589 931.2547,-313.7008 931.853,-306.7264"/>
</g>
<!-- Node28->Node8 -->
<g id="edge73" class="edge">
<title>Node28->Node8</title>
-<path fill="none" stroke="#191970" d="M1932.0678,-363.9442C1941.5056,-361.8162 1951.5853,-359.6923 1961,-358 2111.3863,-330.9682 2236.8789,-398.632 2289,-255 2292.0321,-246.6443 2294.3716,-242.0822 2289,-235 2275.0545,-216.6134 2111.2626,-185.0762 2089,-179 2046.1578,-167.3069 1997.242,-154.3354 1962.3812,-145.156"/>
-<polygon fill="#191970" stroke="#191970" points="1963.0492,-141.7127 1952.4878,-142.5533 1961.2683,-148.4824 1963.0492,-141.7127"/>
+<path fill="none" stroke="#191970" d="M757.163,-363.9032C704.0213,-352.2595 636.1251,-329.2804 665,-291 780.3365,-138.0949 898.5381,-238.2492 1086,-199 1120.6074,-191.7542 1128.382,-186.1951 1163,-179 1238.8934,-163.2261 1326.8489,-150.097 1386.4361,-141.9567"/>
+<polygon fill="#191970" stroke="#191970" points="1387.0458,-145.4062 1396.485,-140.5945 1386.1055,-138.4696 1387.0458,-145.4062"/>
</g>
<!-- Node28->Node9 -->
<g id="edge48" class="edge">
<title>Node28->Node9</title>
-<path fill="none" stroke="#191970" d="M1830.3952,-370.4797C1658.5598,-362.0059 1190.3499,-337.6098 1161,-322 1117.6707,-298.9551 1129.0094,-268.3439 1093,-235 1026.7603,-173.6637 935.2922,-119.3684 884.6171,-91.4354"/>
-<polygon fill="#191970" stroke="#191970" points="886.0984,-88.2564 875.645,-86.5308 882.7407,-94.3986 886.0984,-88.2564"/>
+<path fill="none" stroke="#191970" d="M750.4261,-370.4261C619.9087,-363.5737 327.5082,-345.7597 289,-322 208.9325,-272.598 202.1404,-233.3772 176,-143 171.6031,-127.7981 174.8475,-110.3781 179.3146,-96.5386"/>
+<polygon fill="#191970" stroke="#191970" points="182.7455,-97.3437 182.8627,-86.7496 176.1645,-94.9584 182.7455,-97.3437"/>
</g>
<!-- Node28->Node16 -->
<g id="edge81" class="edge">
<title>Node28->Node16</title>
-<path fill="none" stroke="#191970" d="M1930.2166,-363.9385C1940.1602,-361.6803 1950.9397,-359.4969 1961,-358 2072.0626,-341.4746 2362.8595,-368.6403 2465,-322 2559.5813,-278.8115 2650.8951,-205.7953 2588,-123 2569.6342,-98.8233 2483.7573,-82.9799 2435.7297,-75.8499"/>
-<polygon fill="#191970" stroke="#191970" points="2436.2074,-72.3827 2425.8106,-74.4213 2435.2095,-79.3112 2436.2074,-72.3827"/>
+<path fill="none" stroke="#191970" d="M750.3389,-370.0325C627.2078,-362.7597 362.6093,-344.6913 328,-322 292.7138,-298.865 268.1505,-271.0963 290,-235 349.9678,-135.9307 734.3915,-87.269 849.4874,-74.799"/>
+<polygon fill="#191970" stroke="#191970" points="849.8883,-78.2762 859.4612,-73.7365 849.1467,-71.3156 849.8883,-78.2762"/>
</g>
<!-- Node28->Node17 -->
<g id="edge79" class="edge">
<title>Node28->Node17</title>
-<path fill="none" stroke="#191970" d="M1930.2212,-363.9691C1940.1648,-361.7107 1950.9431,-359.5195 1961,-358 2065.4512,-342.2182 2343.4565,-374.7153 2435,-322 2466.3016,-303.975 2470.2971,-289.4984 2481,-255 2521.9614,-122.9698 2274.5863,-84.3521 2180.3495,-74.5104"/>
-<polygon fill="#191970" stroke="#191970" points="2180.459,-71.0041 2170.1616,-73.4971 2179.7661,-77.9698 2180.459,-71.0041"/>
+<path fill="none" stroke="#191970" d="M750.4801,-370.2136C640.509,-363.789 421.8661,-347.8868 399,-322 389.8788,-311.6738 394.9534,-304.1701 399,-291 407.8341,-262.2485 412.028,-251.7657 437,-235 562.1897,-150.9502 1041.6604,-90.9149 1176.5218,-75.4331"/>
+<polygon fill="#191970" stroke="#191970" points="1177.1664,-78.8824 1186.7063,-74.2737 1176.3746,-71.9273 1177.1664,-78.8824"/>
</g>
<!-- Node28->Node18 -->
<g id="edge82" class="edge">
<title>Node28->Node18</title>
-<path fill="none" stroke="#191970" d="M1890.6121,-363.9304C1887.101,-347.5418 1882.4272,-313.6805 1897,-291 1903.7048,-280.5648 1930.5637,-267.4791 1953.0387,-257.9352"/>
-<polygon fill="#191970" stroke="#191970" points="1954.4356,-261.1449 1962.3309,-254.0798 1951.753,-254.6793 1954.4356,-261.1449"/>
+<path fill="none" stroke="#191970" d="M852.4289,-363.9456C862.9573,-361.6873 874.365,-359.5022 885,-358 1116.1435,-325.3518 1181.6005,-374.7512 1409,-322 1444.536,-313.7565 1450.2795,-302.1915 1485,-291 1539.1538,-273.5446 1603.8334,-259.3056 1642.3263,-251.5031"/>
+<polygon fill="#191970" stroke="#191970" points="1643.3395,-254.8698 1652.4584,-249.4758 1641.966,-248.0059 1643.3395,-254.8698"/>
</g>
<!-- Node28->Node26 -->
<g id="edge72" class="edge">
<title>Node28->Node26</title>
-<path fill="none" stroke="#191970" d="M1830.2789,-369.9686C1629.8499,-358.684 1005.0123,-323.5041 782.5957,-310.9814"/>
-<polygon fill="#191970" stroke="#191970" points="782.4929,-307.4702 772.3119,-310.4024 782.0993,-314.4591 782.4929,-307.4702"/>
+<path fill="none" stroke="#191970" d="M850.8929,-363.9771C861.8303,-361.6008 873.829,-359.348 885,-358 1449.5739,-289.8722 1596.4912,-358.3156 2164,-322 2188.5597,-320.4284 2215.2445,-318.0155 2239.4403,-315.5581"/>
+<polygon fill="#191970" stroke="#191970" points="2240.0565,-319.0132 2249.6441,-314.5045 2239.3374,-312.0502 2240.0565,-319.0132"/>
</g>
<!-- Node29 -->
<g id="node30" class="node">
<title>Node29</title>
<g id="a_node30"><a xlink:href="optional_8h.html" target="_top" xlink:title="Runtime Optional container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1610,-291.5 1610,-321.5 1736,-321.5 1736,-291.5 1610,-291.5"/>
-<text text-anchor="start" x="1618" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
-<text text-anchor="middle" x="1673" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/optional.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="674,-291.5 674,-321.5 800,-321.5 800,-291.5 674,-291.5"/>
+<text text-anchor="start" x="682" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
+<text text-anchor="middle" x="737" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/optional.h</text>
</a>
</g>
</g>
<!-- Node28->Node29 -->
<g id="edge49" class="edge">
<title>Node28->Node29</title>
-<path fill="none" stroke="#191970" d="M1861.4794,-363.9005C1827.8211,-353.6501 1773.708,-337.1702 1732.0794,-324.4924"/>
-<polygon fill="#191970" stroke="#191970" points="1732.9111,-321.087 1722.3252,-321.5218 1730.8717,-327.7834 1732.9111,-321.087"/>
+<path fill="none" stroke="#191970" d="M802.1111,-363.9005C791.6121,-354.6448 775.3518,-340.3101 761.7075,-328.2816"/>
+<polygon fill="#191970" stroke="#191970" points="763.8554,-325.5093 754.0396,-321.5218 759.2264,-330.7602 763.8554,-325.5093"/>
</g>
<!-- Node30 -->
<g id="node31" class="node">
<title>Node30</title>
<g id="a_node31"><a xlink:href="shape__tuple_8h.html" target="_top" xlink:title="Runtime ShapeTuple container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1906,-291.5 1906,-321.5 2032,-321.5 2032,-291.5 1906,-291.5"/>
-<text text-anchor="start" x="1914" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
-<text text-anchor="middle" x="1969" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/shape_tuple.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1092,-291.5 1092,-321.5 1218,-321.5 1218,-291.5 1092,-291.5"/>
+<text text-anchor="start" x="1100" y="-309.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
+<text text-anchor="middle" x="1155" y="-298.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/shape_tuple.h</text>
</a>
</g>
</g>
<!-- Node28->Node30 -->
<g id="edge52" class="edge">
<title>Node28->Node30</title>
-<path fill="none" stroke="#191970" d="M1903.8889,-363.9005C1914.3879,-354.6448 1930.6482,-340.3101 1944.2925,-328.2816"/>
-<polygon fill="#191970" stroke="#191970" points="1946.7736,-330.7602 1951.9604,-321.5218 1942.1446,-325.5093 1946.7736,-330.7602"/>
+<path fill="none" stroke="#191970" d="M856.3694,-363.9906C865.7991,-361.9714 875.7272,-359.8822 885,-358 951.5896,-344.4836 1027.632,-330.0709 1082.0117,-319.9403"/>
+<polygon fill="#191970" stroke="#191970" points="1082.7151,-323.3696 1091.9064,-318.0997 1081.4349,-316.4876 1082.7151,-323.3696"/>
</g>
<!-- Node28->Node31 -->
<g id="edge56" class="edge">
<title>Node28->Node31</title>
-<path fill="none" stroke="#191970" d="M1830.4708,-370.5889C1724.4969,-365.0932 1502.7629,-351.1671 1306.0735,-322.045"/>
-<polygon fill="#191970" stroke="#191970" points="1306.5421,-318.5763 1296.1341,-320.5571 1305.5057,-325.4991 1306.5421,-318.5763"/>
+<path fill="none" stroke="#191970" d="M853.5995,-363.9644C863.8094,-361.7861 874.7728,-359.6343 885,-358 1035.9249,-333.8823 1075.3043,-340.6658 1227,-322 1238.8887,-320.5371 1251.4834,-318.8717 1263.7419,-317.1882"/>
+<polygon fill="#191970" stroke="#191970" points="1264.5128,-320.6148 1273.9369,-315.7738 1263.5508,-313.6812 1264.5128,-320.6148"/>
</g>
<!-- Node35 -->
<g id="node36" class="node">
<title>Node35</title>
<g id="a_node36"><a xlink:href="serializer_8h.html" target="_top" xlink:title="Serializer extension to support TVM data types Include this file to enable serialization of DLDataTyp...">
-<polygon fill="#ffffff" stroke="#000000" points="2292,-297 2292,-316 2426,-316 2426,-297 2292,-297"/>
-<text text-anchor="middle" x="2359" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/serializer.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="408,-297 408,-316 542,-316 542,-297 408,-297"/>
+<text text-anchor="middle" x="475" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/serializer.h</text>
</a>
</g>
</g>
<!-- Node28->Node35 -->
<g id="edge74" class="edge">
<title>Node28->Node35</title>
-<path fill="none" stroke="#191970" d="M1928.5339,-363.9404C1938.9552,-361.8126 1950.3607,-359.6897 1961,-358 2097.1149,-336.3829 2132.7832,-342.9654 2269,-322 2276.9478,-320.7767 2285.307,-319.3442 2293.6095,-317.853"/>
-<polygon fill="#191970" stroke="#191970" points="2294.4683,-321.2542 2303.6772,-316.0152 2293.2112,-314.368 2294.4683,-321.2542"/>
+<path fill="none" stroke="#191970" d="M760.045,-363.9717C696.7049,-351.8833 591.4831,-331.0497 528.0875,-318.0638"/>
+<polygon fill="#191970" stroke="#191970" points="528.7846,-314.634 518.2848,-316.0496 527.3757,-321.4908 528.7846,-314.634"/>
</g>
<!-- Node28->Node38 -->
<g id="edge80" class="edge">
<title>Node28->Node38</title>
-<path fill="none" stroke="#191970" d="M1935.1897,-363.932C1946.449,-361.6739 1958.6432,-359.4922 1970,-358 2223.4342,-324.7017 2293.0164,-370.4242 2544,-322 2547.9677,-321.2345 2552.0639,-320.2271 2556.1075,-319.0925"/>
-<polygon fill="#191970" stroke="#191970" points="2557.4123,-322.3533 2565.947,-316.0757 2555.3603,-315.6608 2557.4123,-322.3533"/>
+<path fill="none" stroke="#191970" d="M823.8889,-363.9005C835.9662,-353.2535 855.6671,-335.8856 870.24,-323.0385"/>
+<polygon fill="#191970" stroke="#191970" points="872.9445,-325.3201 878.1313,-316.0817 868.3154,-320.0692 872.9445,-325.3201"/>
</g>
<!-- Node29->Node16 -->
<g id="edge50" class="edge">
<title>Node29->Node16</title>
-<path fill="none" stroke="#191970" d="M1736.3367,-292.0319C1737.9064,-291.6833 1739.4628,-291.3389 1741,-291 1815.9954,-274.4659 1837.0601,-279.0303 1910,-255 1930.4023,-248.2784 1933.2838,-240.6809 1954,-235 2096.284,-195.9824 2152.0798,-265.0618 2284,-199 2333.8644,-174.0294 2373.5744,-118.9937 2391.9309,-90.0787"/>
-<polygon fill="#191970" stroke="#191970" points="2395.042,-91.7021 2397.3213,-81.3553 2389.0872,-88.0224 2395.042,-91.7021"/>
+<path fill="none" stroke="#191970" d="M742.7455,-291.312C755.4975,-258.6693 788.3573,-180.1554 830,-123 839.4205,-110.0702 852.1527,-97.4015 862.6605,-87.8646"/>
+<polygon fill="#191970" stroke="#191970" points="864.9936,-90.4736 870.1705,-81.2297 860.359,-85.2277 864.9936,-90.4736"/>
</g>
<!-- Node29->Node22 -->
<g id="edge51" class="edge">
<title>Node29->Node22</title>
-<path fill="none" stroke="#191970" d="M1695.4944,-291.2977C1709.3967,-281.9022 1727.1876,-269.8787 1741.1546,-260.4395"/>
-<polygon fill="#191970" stroke="#191970" points="1743.376,-263.1626 1749.7015,-254.6633 1739.4564,-257.3629 1743.376,-263.1626"/>
+<path fill="none" stroke="#191970" d="M800.0476,-297.0977C815.3932,-294.9525 831.7717,-292.7824 847,-291 891.8256,-285.7535 1194.8743,-258.1588 1302.3194,-248.4128"/>
+<polygon fill="#191970" stroke="#191970" points="1302.7425,-251.8889 1312.3856,-247.5001 1302.1103,-244.9175 1302.7425,-251.8889"/>
</g>
<!-- Node30->Node16 -->
<g id="edge53" class="edge">
<title>Node30->Node16</title>
-<path fill="none" stroke="#191970" d="M2007.1329,-291.3832C2048.7797,-275.3792 2117.7323,-250.3539 2179,-235 2226.8405,-223.011 2365.3485,-235.9638 2398,-199 2424.2826,-169.2463 2416.2176,-118.6677 2409.0145,-91.0019"/>
-<polygon fill="#191970" stroke="#191970" points="2412.3602,-89.9704 2406.274,-81.2981 2405.6237,-91.873 2412.3602,-89.9704"/>
+<path fill="none" stroke="#191970" d="M1110.3444,-291.4402C1082.8269,-281.7957 1046.9981,-268.5782 1016,-255 1009.7187,-252.2486 910.1116,-204.4881 906,-199 882.1297,-167.1384 879.7803,-118.4948 880.6133,-91.431"/>
+<polygon fill="#191970" stroke="#191970" points="884.1177,-91.4247 881.1116,-81.2653 877.1261,-91.0819 884.1177,-91.4247"/>
</g>
<!-- Node30->Node18 -->
<g id="edge54" class="edge">
<title>Node30->Node18</title>
-<path fill="none" stroke="#191970" d="M1973.2022,-291.2977C1975.4391,-283.2055 1978.2149,-273.1639 1980.609,-264.5029"/>
-<polygon fill="#191970" stroke="#191970" points="1983.9872,-265.418 1983.2781,-254.8469 1977.2402,-263.5529 1983.9872,-265.418"/>
+<path fill="none" stroke="#191970" d="M1218.072,-297.2976C1233.4168,-295.1458 1249.7887,-292.9225 1265,-291 1404.9137,-273.3164 1572.0756,-255.6892 1642.309,-248.4418"/>
+<polygon fill="#191970" stroke="#191970" points="1642.753,-251.9147 1652.3421,-247.4092 1642.0363,-244.9515 1642.753,-251.9147"/>
</g>
<!-- Node30->Node22 -->
<g id="edge55" class="edge">
<title>Node30->Node22</title>
-<path fill="none" stroke="#191970" d="M1918.8514,-291.4554C1882.5426,-280.5628 1834.2536,-266.0761 1801.3149,-256.1945"/>
-<polygon fill="#191970" stroke="#191970" points="1802.1996,-252.8058 1791.6156,-253.2847 1800.1881,-259.5106 1802.1996,-252.8058"/>
+<path fill="none" stroke="#191970" d="M1200.2561,-291.4554C1231.779,-280.9762 1273.3083,-267.1705 1302.8827,-257.339"/>
+<polygon fill="#191970" stroke="#191970" points="1304.037,-260.6437 1312.4223,-254.1677 1301.8288,-254.0011 1304.037,-260.6437"/>
</g>
<!-- Node31->Node7 -->
<g id="edge57" class="edge">
<title>Node31->Node7</title>
-<path fill="none" stroke="#191970" d="M1296.1655,-292.6336C1299.1484,-292.0632 1302.1046,-291.5153 1305,-291 1417.6861,-270.943 1448.6437,-281.4618 1560,-255 1618.0944,-241.195 1683.4144,-217.2986 1721.7859,-202.3623"/>
-<polygon fill="#191970" stroke="#191970" points="1723.3738,-205.499 1731.4028,-198.5863 1720.8154,-198.9833 1723.3738,-205.499"/>
+<path fill="none" stroke="#191970" d="M1349.5503,-291.1389C1367.3967,-269.2957 1400.0149,-229.3724 1418.7492,-206.4424"/>
+<polygon fill="#191970" stroke="#191970" points="1421.5731,-208.5179 1425.1897,-198.5595 1416.1523,-204.089 1421.5731,-208.5179"/>
</g>
<!-- Node31->Node8 -->
<g id="edge61" class="edge">
<title>Node31->Node8</title>
-<path fill="none" stroke="#191970" d="M1296.214,-292.4739C1360.9477,-278.0255 1452.6052,-257.2947 1456,-255 1490.1171,-231.9386 1473.6201,-200.0729 1509,-179 1537.2606,-162.1674 1737.2892,-145.6365 1846.0218,-137.7647"/>
-<polygon fill="#191970" stroke="#191970" points="1846.5269,-141.2375 1856.2505,-137.0302 1846.0254,-134.2555 1846.5269,-141.2375"/>
+<path fill="none" stroke="#191970" d="M1363.5158,-291.4121C1402.6637,-268.5122 1472.7427,-224.9123 1486,-199 1494.3555,-182.6685 1483.4726,-163.5398 1472.549,-150.0927"/>
+<polygon fill="#191970" stroke="#191970" points="1475.14,-147.7388 1465.914,-142.53 1469.878,-152.3553 1475.14,-147.7388"/>
</g>
<!-- Node31->Node13 -->
<g id="edge59" class="edge">
<title>Node31->Node13</title>
-<path fill="none" stroke="#191970" d="M1295.949,-291.4691C1346.9521,-278.955 1412.0877,-262.0809 1423,-255 1459.5678,-231.2714 1448.9021,-203.4375 1485,-179 1568.0401,-122.7836 1681.9769,-94.7854 1755.3038,-81.73"/>
-<polygon fill="#191970" stroke="#191970" points="1756.1178,-85.1412 1765.3736,-79.9856 1754.9229,-78.2439 1756.1178,-85.1412"/>
+<path fill="none" stroke="#191970" d="M1380.2089,-291.4826C1402.225,-282.7058 1428.7574,-270.3269 1450,-255 1477.5434,-235.1269 1483.9196,-227.752 1502,-199 1523.7222,-164.4567 1535.3776,-117.5228 1540.5753,-91.306"/>
+<polygon fill="#191970" stroke="#191970" points="1544.0793,-91.6074 1542.4702,-81.1354 1537.1977,-90.3252 1544.0793,-91.6074"/>
</g>
<!-- Node31->Node14 -->
<g id="edge67" class="edge">
<title>Node31->Node14</title>
-<path fill="none" stroke="#191970" d="M1209.843,-291.293C1182.9538,-274.4583 1136.6875,-247.9889 1093,-235 947.7691,-191.8209 902.8121,-227.4854 754,-199 630.7191,-175.4018 603.8895,-152.967 482,-123 407.3125,-104.6378 318.7164,-86.768 272.2045,-77.6845"/>
-<polygon fill="#191970" stroke="#191970" points="272.612,-74.1983 262.1279,-75.7257 271.2762,-81.0697 272.612,-74.1983"/>
+<path fill="none" stroke="#191970" d="M1400.1788,-291.9461C1444.6152,-281.6847 1505.4633,-267.5831 1559,-255 1596.3589,-246.2193 1605.3627,-242.4983 1643,-235 2027.7712,-158.344 2500.0336,-92.8909 2626.5554,-75.8273"/>
+<polygon fill="#191970" stroke="#191970" points="2627.2821,-79.2612 2636.7263,-74.4596 2626.3491,-72.3237 2627.2821,-79.2612"/>
</g>
<!-- Node31->Node15 -->
<g id="edge70" class="edge">
<title>Node31->Node15</title>
-<path fill="none" stroke="#191970" d="M1227.719,-291.2108C1221.3952,-275.0594 1209.3074,-249.9089 1191,-235 1113.8534,-172.1744 816.8826,-104.8129 702.9009,-80.7168"/>
-<polygon fill="#191970" stroke="#191970" points="703.2793,-77.2199 692.7732,-78.588 701.8394,-84.0702 703.2793,-77.2199"/>
+<path fill="none" stroke="#191970" d="M1384.9489,-291.4817C1412.4411,-282.2282 1447.2843,-269.3974 1477,-255 1535.1708,-226.8161 1540.3121,-201.2519 1601,-179 1722.4365,-134.4739 2118.176,-90.2217 2252.7367,-76.1193"/>
+<polygon fill="#191970" stroke="#191970" points="2253.5005,-79.5587 2263.0836,-75.0405 2252.7746,-72.5964 2253.5005,-79.5587"/>
</g>
<!-- Node31->Node16 -->
<g id="edge69" class="edge">
<title>Node31->Node16</title>
-<path fill="none" stroke="#191970" d="M1296.1262,-292.389C1299.1196,-291.884 1302.0888,-291.4169 1305,-291 1523.7919,-259.6657 1590.4209,-322.143 1801,-255 1817.3395,-249.7901 1818.2874,-241.8729 1834,-235 2029.2304,-149.604 2281.4302,-95.291 2370.2816,-77.7237"/>
-<polygon fill="#191970" stroke="#191970" points="2371.1514,-81.1199 2380.2926,-75.7639 2369.8065,-74.2503 2371.1514,-81.1199"/>
+<path fill="none" stroke="#191970" d="M1273.8005,-297.0569C1203.2844,-286.1576 1093.864,-268.0595 1054,-255 1047.247,-252.7877 943.9418,-204.1064 939,-199 923.2446,-182.7199 900.1178,-122.267 888.7771,-90.8071"/>
+<polygon fill="#191970" stroke="#191970" points="892.0631,-89.6012 885.4107,-81.3552 885.4689,-91.9499 892.0631,-89.6012"/>
</g>
<!-- Node31->Node18 -->
<g id="edge71" class="edge">
<title>Node31->Node18</title>
-<path fill="none" stroke="#191970" d="M1296.1236,-292.3709C1299.1177,-291.8708 1302.0878,-291.4096 1305,-291 1313.1135,-289.8588 1815.9064,-256.3284 1952.2258,-247.2486"/>
-<polygon fill="#191970" stroke="#191970" points="1952.7322,-250.7227 1962.4775,-246.5658 1952.2669,-243.7382 1952.7322,-250.7227"/>
+<path fill="none" stroke="#191970" d="M1400.3656,-295.0045C1472.0197,-282.0053 1586.0164,-261.3244 1642.2985,-251.114"/>
+<polygon fill="#191970" stroke="#191970" points="1643.2199,-254.5041 1652.4345,-249.2752 1641.9703,-247.6165 1643.2199,-254.5041"/>
</g>
<!-- Node31->Node20 -->
<g id="edge62" class="edge">
<title>Node31->Node20</title>
-<path fill="none" stroke="#191970" d="M1296.182,-292.7231C1299.1605,-292.1287 1302.1113,-291.5514 1305,-291 1400.5985,-272.7537 1428.145,-284.1539 1521,-255 1562.4947,-241.9718 1607.4289,-218.8213 1634.6458,-203.6923"/>
-<polygon fill="#191970" stroke="#191970" points="1636.5041,-206.6626 1643.4992,-198.7053 1633.0686,-200.5637 1636.5041,-206.6626"/>
+<path fill="none" stroke="#191970" d="M1400.162,-293.1238C1437.0081,-284.337 1483.9676,-271.4253 1524,-255 1558.6998,-240.7626 1596.0543,-218.5852 1619.2306,-203.9265"/>
+<polygon fill="#191970" stroke="#191970" points="1621.1366,-206.8622 1627.6713,-198.5226 1617.3622,-200.9669 1621.1366,-206.8622"/>
</g>
<!-- Node31->Node21 -->
<g id="edge66" class="edge">
<title>Node31->Node21</title>
-<path fill="none" stroke="#191970" d="M1189.4563,-291.4554C1159.6263,-281.149 1120.4834,-267.6249 1092.1298,-257.8286"/>
-<polygon fill="#191970" stroke="#191970" points="1093.1784,-254.4879 1082.5837,-254.5303 1090.8924,-261.1041 1093.1784,-254.4879"/>
+<path fill="none" stroke="#191970" d="M1400.0378,-298.523C1501.8833,-285.6351 1699.6806,-260.6052 1783.711,-249.9718"/>
+<polygon fill="#191970" stroke="#191970" points="1784.3707,-253.4163 1793.8522,-248.6885 1783.4919,-246.4717 1784.3707,-253.4163"/>
</g>
<!-- Node31->Node22 -->
<g id="edge58" class="edge">
<title>Node31->Node22</title>
-<path fill="none" stroke="#191970" d="M1296.1411,-292.4881C1299.1305,-291.9566 1302.0948,-291.4568 1305,-291 1460.2771,-266.5842 1647.2556,-252.568 1726.056,-247.3566"/>
-<polygon fill="#191970" stroke="#191970" points="1726.6914,-250.8227 1736.4425,-246.6795 1726.236,-243.8375 1726.6914,-250.8227"/>
+<path fill="none" stroke="#191970" d="M1337.7416,-291.2977C1338.1276,-283.3834 1338.6047,-273.6043 1339.0207,-265.0759"/>
+<polygon fill="#191970" stroke="#191970" points="1342.5282,-265.0056 1339.5197,-254.8469 1335.5365,-264.6645 1342.5282,-265.0056"/>
</g>
<!-- Node31->Node23 -->
<g id="edge60" class="edge">
<title>Node31->Node23</title>
-<path fill="none" stroke="#191970" d="M1223.0956,-291.3087C1211.2274,-274.4885 1189.6924,-248.0323 1164,-235 1119.4677,-212.4113 986.6914,-199.5921 901.7031,-193.4928"/>
-<polygon fill="#191970" stroke="#191970" points="901.8054,-189.9915 891.5847,-192.7821 901.3149,-196.9743 901.8054,-189.9915"/>
+<path fill="none" stroke="#191970" d="M1273.8967,-297.334C1210.9481,-287.5115 1119.1208,-271.1444 1087,-255 1063.3803,-243.1284 1041.4704,-221.7508 1027.7179,-206.5434"/>
+<polygon fill="#191970" stroke="#191970" points="1030.1808,-204.0433 1020.9578,-198.8291 1024.9161,-208.6568 1030.1808,-204.0433"/>
</g>
<!-- Node31->Node25 -->
<g id="edge65" class="edge">
<title>Node31->Node25</title>
-<path fill="none" stroke="#191970" d="M1296.1983,-292.8058C1299.1724,-292.1893 1302.1178,-291.5846 1305,-291 1388.8035,-274.0007 1418.5767,-295.2906 1494,-255 1515.5814,-243.4714 1534.4686,-222.335 1546.2416,-207.0825"/>
-<polygon fill="#191970" stroke="#191970" points="1549.1665,-209.0143 1552.3179,-198.899 1543.5463,-204.8413 1549.1665,-209.0143"/>
+<path fill="none" stroke="#191970" d="M1321.1717,-291.3795C1298.0874,-269.3274 1255.3354,-228.4871 1231.396,-205.6181"/>
+<polygon fill="#191970" stroke="#191970" points="1233.7028,-202.9814 1224.0542,-198.6046 1228.8674,-208.043 1233.7028,-202.9814"/>
</g>
<!-- Node32 -->
<g id="node33" class="node">
<title>Node32</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1360.5,-235.5 1360.5,-254.5 1413.5,-254.5 1413.5,-235.5 1360.5,-235.5"/>
-<text text-anchor="middle" x="1387" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstddef</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1096.5,-235.5 1096.5,-254.5 1149.5,-254.5 1149.5,-235.5 1096.5,-235.5"/>
+<text text-anchor="middle" x="1123" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstddef</text>
</g>
<!-- Node31->Node32 -->
<g id="edge63" class="edge">
<title>Node31->Node32</title>
-<path fill="none" stroke="#191970" d="M1270.6726,-291.4554C1296.0298,-281.329 1329.1652,-268.0964 1353.5826,-258.3453"/>
-<polygon fill="#191970" stroke="#191970" points="1355.1467,-261.4895 1363.1355,-254.5303 1352.5505,-254.9887 1355.1467,-261.4895"/>
+<path fill="none" stroke="#191970" d="M1284.6497,-291.4554C1245.7952,-280.2893 1193.7985,-265.3463 1159.3937,-255.4589"/>
+<polygon fill="#191970" stroke="#191970" points="1160.3267,-252.0855 1149.7489,-252.6872 1158.3932,-258.8132 1160.3267,-252.0855"/>
</g>
<!-- Node33 -->
<g id="node34" class="node">
<title>Node33</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="1292,-235.5 1292,-254.5 1342,-254.5 1342,-235.5 1292,-235.5"/>
-<text text-anchor="middle" x="1317" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstring</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1168,-235.5 1168,-254.5 1218,-254.5 1218,-235.5 1168,-235.5"/>
+<text text-anchor="middle" x="1193" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">cstring</text>
</g>
<!-- Node31->Node33 -->
<g id="edge64" class="edge">
<title>Node31->Node33</title>
-<path fill="none" stroke="#191970" d="M1253.7641,-291.2977C1266.3958,-282.0495 1282.5053,-270.2551 1295.3031,-260.8853"/>
-<polygon fill="#191970" stroke="#191970" points="1297.5495,-263.5784 1303.5505,-254.8469 1293.4143,-257.9303 1297.5495,-263.5784"/>
+<path fill="none" stroke="#191970" d="M1301.7736,-291.4554C1278.1684,-281.374 1247.3546,-268.2139 1224.5524,-258.4755"/>
+<polygon fill="#191970" stroke="#191970" points="1225.886,-255.2393 1215.3149,-254.5303 1223.1366,-261.6767 1225.886,-255.2393"/>
</g>
<!-- Node34 -->
<g id="node35" class="node">
<title>Node34</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="914.5,-235.5 914.5,-254.5 1007.5,-254.5 1007.5,-235.5 914.5,-235.5"/>
-<text text-anchor="middle" x="961" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">unordered_map</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="870.5,-235.5 870.5,-254.5 963.5,-254.5 963.5,-235.5 870.5,-235.5"/>
+<text text-anchor="middle" x="917" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">unordered_map</text>
</g>
<!-- Node31->Node34 -->
<g id="edge68" class="edge">
<title>Node31->Node34</title>
-<path fill="none" stroke="#191970" d="M1169.9141,-292.2361C1122.35,-281.4817 1057.8926,-266.9077 1013.2086,-256.8045"/>
-<polygon fill="#191970" stroke="#191970" points="1013.951,-253.3841 1003.4253,-254.5925 1012.4072,-260.2118 1013.951,-253.3841"/>
+<path fill="none" stroke="#191970" d="M1273.667,-297.2262C1192.1915,-285.2959 1051.8846,-264.751 973.7401,-253.3084"/>
+<polygon fill="#191970" stroke="#191970" points="974.003,-249.8096 963.6014,-251.8238 972.9888,-256.7358 974.003,-249.8096"/>
</g>
<!-- Node35->Node9 -->
<g id="edge77" class="edge">
<title>Node35->Node9</title>
-<path fill="none" stroke="#191970" d="M2353.9295,-296.7309C2345.5308,-281.4584 2327.3204,-251.9269 2304,-235 2182.3946,-146.734 2132.5588,-151.3755 1985,-123 1780.3193,-83.64 1141.2204,-74.1819 921.8219,-72.0687"/>
-<polygon fill="#191970" stroke="#191970" points="921.7275,-68.5678 911.695,-71.9735 921.6617,-75.5674 921.7275,-68.5678"/>
+<path fill="none" stroke="#191970" d="M463.4747,-296.9967C421.5023,-262.3879 276.1457,-142.5325 216.1808,-93.0877"/>
+<polygon fill="#191970" stroke="#191970" points="218.2479,-90.2558 208.3059,-86.5943 213.7946,-95.6566 218.2479,-90.2558"/>
</g>
<!-- Node35->Node28 -->
<g id="edge78" class="edge">
<title>Node35->Node28</title>
-<path fill="none" stroke="#191970" d="M2321.1482,-316.0152C2310.1725,-318.1446 2298.1793,-320.2794 2287,-322 2150.7832,-342.9654 2115.1149,-336.3829 1979,-358 1971.5193,-359.1881 1963.6597,-360.5903 1955.8459,-362.0578"/>
-<polygon fill="#191970" stroke="#191970" points="1955.1692,-358.6237 1946.0049,-363.9404 1956.4845,-365.499 1955.1692,-358.6237"/>
+<path fill="none" stroke="#191970" d="M528.0666,-316.0496C591.3255,-328.1234 696.2449,-348.8967 759.6326,-361.8788"/>
+<polygon fill="#191970" stroke="#191970" points="759.3187,-365.3874 769.8186,-363.9717 760.7276,-358.5306 759.3187,-365.3874"/>
</g>
<!-- Node36 -->
<g id="node37" class="node">
<title>Node36</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2218,-235.5 2218,-254.5 2280,-254.5 2280,-235.5 2218,-235.5"/>
-<text text-anchor="middle" x="2249" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/io.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="600,-235.5 600,-254.5 662,-254.5 662,-235.5 600,-235.5"/>
+<text text-anchor="middle" x="631" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/io.h</text>
</g>
<!-- Node35->Node36 -->
<g id="edge75" class="edge">
<title>Node35->Node36</title>
-<path fill="none" stroke="#191970" d="M2341.8125,-296.8906C2323.8918,-286.8713 2295.5905,-271.0483 2274.9477,-259.5071"/>
-<polygon fill="#191970" stroke="#191970" points="2276.5033,-256.367 2266.0668,-254.5419 2273.0872,-262.4769 2276.5033,-256.367"/>
+<path fill="none" stroke="#191970" d="M499.375,-296.8906C525.7056,-286.5103 567.838,-269.9004 597.3229,-258.2766"/>
+<polygon fill="#191970" stroke="#191970" points="598.7767,-261.4657 606.7962,-254.5419 596.2093,-254.9534 598.7767,-261.4657"/>
</g>
<!-- Node37 -->
<g id="node38" class="node">
<title>Node37</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2374.5,-235.5 2374.5,-254.5 2471.5,-254.5 2471.5,-235.5 2374.5,-235.5"/>
-<text text-anchor="middle" x="2423" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/serializer.h</text>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="446.5,-235.5 446.5,-254.5 543.5,-254.5 543.5,-235.5 446.5,-235.5"/>
+<text text-anchor="middle" x="495" y="-242.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/serializer.h</text>
</g>
<!-- Node35->Node37 -->
<g id="edge76" class="edge">
<title>Node35->Node37</title>
-<path fill="none" stroke="#191970" d="M2369,-296.8906C2378.7119,-287.5581 2393.6639,-273.1902 2405.3741,-261.9374"/>
-<polygon fill="#191970" stroke="#191970" points="2408.0736,-264.1974 2412.859,-254.7449 2403.2234,-259.15 2408.0736,-264.1974"/>
+<path fill="none" stroke="#191970" d="M478.125,-296.8906C480.9265,-288.276 485.1233,-275.3708 488.6306,-264.5858"/>
+<polygon fill="#191970" stroke="#191970" points="492.0667,-265.3371 491.8309,-254.7449 485.4098,-263.1722 492.0667,-265.3371"/>
</g>
<!-- Node39->Node8 -->
-<g id="edge113" class="edge">
+<g id="edge114" class="edge">
<title>Node39->Node8</title>
-<path fill="none" stroke="#191970" d="M1126.2834,-425.8167C1127.8694,-425.5278 1129.4438,-425.2546 1131,-425 1226.0685,-409.4441 1485.848,-446.6851 1563,-389 1600.4138,-361.0264 1564.9559,-320.7177 1601,-291 1670.6862,-233.545 1719.318,-293.5395 1801,-255 1849.273,-232.2237 1887.3966,-179.4772 1905.1415,-151.3842"/>
-<polygon fill="#191970" stroke="#191970" points="1908.3059,-152.9192 1910.5514,-142.565 1902.339,-149.2589 1908.3059,-152.9192"/>
+<path fill="none" stroke="#191970" d="M1794.3391,-425.3213C1812.3259,-416.7399 1833.0411,-404.5808 1848,-389 1896.353,-338.6367 1913.9065,-295.4649 1879,-235 1856.1958,-195.4986 1835.7447,-194.915 1793,-179 1768.5309,-169.8895 1617.0099,-151.4102 1525.6766,-140.8567"/>
+<polygon fill="#191970" stroke="#191970" points="1526.0035,-137.3713 1515.6687,-139.7042 1525.2026,-144.3254 1526.0035,-137.3713"/>
</g>
<!-- Node39->Node9 -->
<g id="edge91" class="edge">
<title>Node39->Node9</title>
-<path fill="none" stroke="#191970" d="M1009.9088,-431.4364C956.8794,-422.2273 883.3181,-406.8118 860,-389 818.4583,-357.2679 738.0469,-228.7809 754,-179 765.3853,-143.4726 795.7554,-112.4171 818.7177,-92.9413"/>
-<polygon fill="#191970" stroke="#191970" points="820.9902,-95.6041 826.4876,-86.5471 816.542,-90.199 820.9902,-95.6041"/>
+<path fill="none" stroke="#191970" d="M1697.6878,-439.7536C1452.118,-436.4152 508.848,-421.5682 380,-389 249.6651,-356.056 114,-379.434 114,-245 114,-245 114,-245 114,-189 114,-150.9686 142.4185,-115.5128 164.4999,-93.7037"/>
+<polygon fill="#191970" stroke="#191970" points="167.1577,-96.006 171.9891,-86.5768 162.3321,-90.9351 167.1577,-96.006"/>
</g>
<!-- Node39->Node13 -->
<g id="edge100" class="edge">
<title>Node39->Node13</title>
-<path fill="none" stroke="#191970" d="M1126.0789,-433.425C1152.293,-430.4441 1183.6698,-427.1698 1212,-425 1290.9204,-418.9556 2580.0588,-380.8405 2633,-322 2736.7174,-206.7253 2117.2727,-110.6652 1900.7107,-80.9824"/>
-<polygon fill="#191970" stroke="#191970" points="1901.0882,-77.5016 1890.7072,-79.6193 1900.1431,-84.4375 1901.0882,-77.5016"/>
+<path fill="none" stroke="#191970" d="M1812.3451,-425.4458C1834.2154,-417.4222 1858.2227,-405.671 1876,-389 1926.4329,-341.7055 1933.2719,-294.4654 1898,-235 1839.1164,-135.7272 1701.6424,-96.1641 1616.3816,-80.8106"/>
+<polygon fill="#191970" stroke="#191970" points="1616.9588,-77.3586 1606.5079,-79.0986 1615.7629,-84.2556 1616.9588,-77.3586"/>
</g>
<!-- Node39->Node14 -->
-<g id="edge117" class="edge">
+<g id="edge118" class="edge">
<title>Node39->Node14</title>
-<path fill="none" stroke="#191970" d="M1009.9651,-436.3323C904.8333,-428.3825 687.0029,-409.9651 614,-389 547.6993,-369.9596 532.7606,-358.132 474,-322 373.5064,-260.2063 344.9687,-240.1217 278,-143 266.3294,-126.0746 255.5913,-105.1523 248.5291,-90.3238"/>
-<polygon fill="#191970" stroke="#191970" points="251.6545,-88.7441 244.261,-81.1554 245.3084,-91.6984 251.6545,-88.7441"/>
+<path fill="none" stroke="#191970" d="M1814.0174,-434.957C1941.5436,-422.717 2236.6626,-394.0633 2258,-389 2339.437,-369.6753 2361.4247,-363.6925 2434,-322 2534.8797,-264.0474 2562.1151,-240.881 2625,-143 2635.7613,-126.2499 2645.2294,-105.5652 2651.4228,-90.7603"/>
+<polygon fill="#191970" stroke="#191970" points="2654.802,-91.7415 2655.3254,-81.1596 2648.3173,-89.1055 2654.802,-91.7415"/>
</g>
<!-- Node39->Node15 -->
-<g id="edge119" class="edge">
+<g id="edge120" class="edge">
<title>Node39->Node15</title>
-<path fill="none" stroke="#191970" d="M1009.7255,-434.136C937.1033,-425.5228 817.7076,-408.9188 778,-389 760.0132,-379.9771 762.7303,-367.517 745,-358 683.4494,-324.9618 635.3722,-376.7618 592,-322 568.4037,-292.2072 497.4909,-355.4843 606,-123 612.5947,-108.8706 624.8162,-96.4532 635.7544,-87.3532"/>
-<polygon fill="#191970" stroke="#191970" points="638.0534,-89.998 643.7543,-81.0677 633.7287,-84.4937 638.0534,-89.998"/>
+<path fill="none" stroke="#191970" d="M1808.7917,-425.4493C1918.322,-394.1852 2160.4467,-324.8642 2164,-322 2242.0308,-259.102 2280.6149,-138.2203 2293.2099,-91.0397"/>
+<polygon fill="#191970" stroke="#191970" points="2296.6442,-91.738 2295.745,-81.1814 2289.8647,-89.9947 2296.6442,-91.738"/>
</g>
<!-- Node39->Node16 -->
-<g id="edge120" class="edge">
+<g id="edge121" class="edge">
<title>Node39->Node16</title>
-<path fill="none" stroke="#191970" d="M1126.0683,-433.2811C1152.2803,-430.2723 1183.6584,-427.016 1212,-425 1372.0045,-413.6183 2548.218,-444.3122 2652,-322 2692.431,-274.3501 2770.2692,-283.9302 2607,-123 2582.9054,-99.2506 2486.8699,-82.9064 2435.6199,-75.6855"/>
-<polygon fill="#191970" stroke="#191970" points="2436.0867,-72.2168 2425.7034,-74.3235 2435.1342,-79.1517 2436.0867,-72.2168"/>
+<path fill="none" stroke="#191970" d="M1697.71,-440.1486C1437.4794,-438.0088 390.2302,-423.1289 289,-322 242.311,-275.3578 275.0481,-217.0073 329,-179 383.1119,-140.88 402.9408,-139.663 467,-123 606.0112,-86.8406 778.0228,-75.8121 849.0015,-72.6675"/>
+<polygon fill="#191970" stroke="#191970" points="849.2711,-76.1594 859.1156,-72.2432 848.9777,-69.1655 849.2711,-76.1594"/>
</g>
<!-- Node39->Node18 -->
-<g id="edge121" class="edge">
+<g id="edge122" class="edge">
<title>Node39->Node18</title>
-<path fill="none" stroke="#191970" d="M1126.0381,-427.3111C1130.7481,-426.4529 1135.4455,-425.667 1140,-425 1425.064,-383.2534 1523.2794,-452.7624 1780,-322 1801.9681,-310.8104 1803.6577,-301.4223 1826,-291 1837.5487,-285.6127 1909.1212,-265.8166 1952.5343,-254.0213"/>
-<polygon fill="#191970" stroke="#191970" points="1953.6512,-257.3449 1962.387,-251.3502 1951.8196,-250.5888 1953.6512,-257.3449"/>
+<path fill="none" stroke="#191970" d="M1758.2599,-425.2451C1760.0483,-407.3985 1760.2569,-377.7804 1746,-358 1724.1461,-327.6793 1689.9866,-353.5828 1670,-322 1659.1774,-304.8982 1663.2026,-281.1955 1668.3293,-264.6207"/>
+<polygon fill="#191970" stroke="#191970" points="1671.7312,-265.4939 1671.7123,-254.8991 1665.1201,-263.1932 1671.7312,-265.4939"/>
</g>
<!-- Node39->Node19 -->
<g id="edge92" class="edge">
<title>Node39->Node19</title>
-<path fill="none" stroke="#191970" d="M1102.8456,-425.389C1159.3276,-400.8952 1270.7827,-352.5619 1332.7734,-325.6792"/>
-<polygon fill="#191970" stroke="#191970" points="1334.4998,-328.7455 1342.2818,-321.5558 1331.7148,-322.3234 1334.4998,-328.7455"/>
+<path fill="none" stroke="#191970" d="M1724.3579,-425.4839C1710.1842,-416.9409 1694.9065,-404.7673 1687,-389 1680.8241,-376.6839 1680.1463,-369.9521 1687,-358 1694.8013,-344.3954 1707.7423,-334.0392 1721.3312,-326.3125"/>
+<polygon fill="#191970" stroke="#191970" points="1723.0755,-329.3499 1730.3071,-321.6069 1719.8253,-323.1502 1723.0755,-329.3499"/>
</g>
<!-- Node39->Node21 -->
-<g id="edge116" class="edge">
+<g id="edge117" class="edge">
<title>Node39->Node21</title>
-<path fill="none" stroke="#191970" d="M1100.9056,-425.4761C1115.3794,-416.9924 1130.9054,-404.8713 1139,-389 1145.2598,-376.7263 1143.3291,-371.08 1139,-358 1125.9918,-318.6964 1093.4301,-281.9394 1072.8806,-261.5618"/>
-<polygon fill="#191970" stroke="#191970" points="1075.2819,-259.0151 1065.6588,-254.5828 1070.4175,-264.0488 1075.2819,-259.0151"/>
+<path fill="none" stroke="#191970" d="M1778.8292,-425.4059C1806.295,-405.5799 1850.5228,-367.847 1865,-322 1869.1487,-308.8617 1869.9821,-303.8455 1865,-291 1860.6142,-279.692 1852.0098,-269.463 1843.7251,-261.5381"/>
+<polygon fill="#191970" stroke="#191970" points="1845.8244,-258.721 1836.028,-254.6859 1841.1699,-263.9494 1845.8244,-258.721"/>
</g>
<!-- Node39->Node26 -->
<g id="edge99" class="edge">
<title>Node39->Node26</title>
-<path fill="none" stroke="#191970" d="M1009.9078,-434.4372C959.6838,-427.7379 885.9331,-414.3567 826,-389 820.5655,-386.7007 759.6641,-345.2677 725.6486,-322.0141"/>
-<polygon fill="#191970" stroke="#191970" points="727.4621,-319.0141 717.2324,-316.2566 723.5097,-324.7916 727.4621,-319.0141"/>
+<path fill="none" stroke="#191970" d="M1814.0666,-433.7028C1886.4653,-424.6238 2005.621,-407.5198 2046,-389 2065.8115,-379.9134 2064.3741,-367.4808 2084,-358 2099.7297,-350.4013 2195.2644,-330.7397 2259.5788,-318.0269"/>
+<polygon fill="#191970" stroke="#191970" points="2260.4811,-321.4165 2269.6157,-316.0492 2259.1277,-314.5486 2260.4811,-321.4165"/>
</g>
<!-- Node39->Node28 -->
-<g id="edge112" class="edge">
+<g id="edge113" class="edge">
<title>Node39->Node28</title>
-<path fill="none" stroke="#191970" d="M1126.1288,-431.4677C1142.8425,-429.0992 1161.1166,-426.7277 1178,-425 1426.9894,-399.5201 1490.5609,-409.617 1740,-389 1766.0751,-386.8448 1794.6073,-384.0683 1819.8946,-381.462"/>
-<polygon fill="#191970" stroke="#191970" points="1820.5623,-384.9115 1830.1469,-380.3964 1819.8386,-377.949 1820.5623,-384.9115"/>
+<path fill="none" stroke="#191970" d="M1697.9078,-438.2274C1529.4761,-431.4539 1042.8758,-410.5321 885,-389 877.5124,-387.9788 869.6445,-386.603 861.971,-385.0907"/>
+<polygon fill="#191970" stroke="#191970" points="862.5448,-381.6354 852.0438,-383.0424 861.1302,-388.491 862.5448,-381.6354"/>
</g>
<!-- Node39->Node38 -->
-<g id="edge114" class="edge">
+<g id="edge115" class="edge">
<title>Node39->Node38</title>
-<path fill="none" stroke="#191970" d="M1126.0811,-433.4531C1152.2955,-430.4776 1183.6721,-427.1998 1212,-425 1379.0216,-412.03 1806.8546,-444.2673 1965,-389 1987.2407,-381.2275 1985.8353,-365.9867 2008,-358 2232.6215,-277.0617 2309.6653,-367.7503 2544,-322 2547.966,-321.2257 2552.0612,-320.2127 2556.1041,-319.075"/>
-<polygon fill="#191970" stroke="#191970" points="2557.4105,-322.3352 2565.9429,-316.0544 2555.356,-315.6434 2557.4105,-322.3352"/>
+<path fill="none" stroke="#191970" d="M1697.9387,-435.6313C1562.8644,-423.5428 1217.6113,-388.2456 936,-322 932.3566,-321.1429 928.5921,-320.1287 924.8534,-319.0357"/>
+<polygon fill="#191970" stroke="#191970" points="925.7058,-315.6355 915.1176,-316.0119 923.6295,-322.3205 925.7058,-315.6355"/>
</g>
<!-- Node40 -->
<g id="node41" class="node">
<title>Node40</title>
<g id="a_node41"><a xlink:href="map_8h.html" target="_top" xlink:title="Runtime Map container types. ">
-<polygon fill="#ffffff" stroke="#000000" points="1428,-358.5 1428,-388.5 1554,-388.5 1554,-358.5 1428,-358.5"/>
-<text text-anchor="start" x="1436" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
-<text text-anchor="middle" x="1491" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/map.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="894,-358.5 894,-388.5 1020,-388.5 1020,-358.5 894,-358.5"/>
+<text text-anchor="start" x="902" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/container</text>
+<text text-anchor="middle" x="957" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/map.h</text>
</a>
</g>
</g>
<!-- Node39->Node40 -->
<g id="edge93" class="edge">
<title>Node39->Node40</title>
-<path fill="none" stroke="#191970" d="M1126.0932,-427.6478C1130.7908,-426.7141 1135.4703,-425.8189 1140,-425 1235.459,-407.7415 1346.3205,-392.2252 1417.4975,-382.8402"/>
-<polygon fill="#191970" stroke="#191970" points="1418.245,-386.2722 1427.7045,-381.5005 1417.3341,-379.3317 1418.245,-386.2722"/>
+<path fill="none" stroke="#191970" d="M1697.8943,-435.6276C1555.1407,-423.657 1187.4373,-392.8233 1030.1444,-379.6335"/>
+<polygon fill="#191970" stroke="#191970" points="1030.3502,-376.1386 1020.0927,-378.7906 1029.7652,-383.1141 1030.3502,-376.1386"/>
</g>
<!-- Node41 -->
<g id="node42" class="node">
<title>Node41</title>
<g id="a_node42"><a xlink:href="runtime_2module_8h.html" target="_top" xlink:title="Runtime container of the functions generated by TVM, This is used to support dynamically link...">
-<polygon fill="#ffffff" stroke="#000000" points="1004,-364 1004,-383 1130,-383 1130,-364 1004,-364"/>
-<text text-anchor="middle" x="1067" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/module.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1438,-364 1438,-383 1564,-383 1564,-364 1438,-364"/>
+<text text-anchor="middle" x="1501" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/module.h</text>
</a>
</g>
</g>
<!-- Node39->Node41 -->
<g id="edge101" class="edge">
<title>Node39->Node41</title>
-<path fill="none" stroke="#191970" d="M1062.0123,-425.2967C1060.7221,-415.7699 1060.3886,-403.3954 1061.0859,-393.1306"/>
-<polygon fill="#191970" stroke="#191970" points="1064.5791,-393.3908 1062.2507,-383.055 1057.6254,-392.5869 1064.5791,-393.3908"/>
-</g>
-<!-- Node39->Node42 -->
-<g id="edge115" class="edge">
-<title>Node39->Node42</title>
-<path fill="none" stroke="#191970" d="M1009.9265,-435.4269C907.6476,-426.4409 688.7139,-406.9762 504,-389 463.3258,-385.0416 416.4846,-380.1477 386.1195,-376.9308"/>
-<polygon fill="#191970" stroke="#191970" points="386.417,-373.4428 376.1033,-375.8672 385.6777,-380.4036 386.417,-373.4428"/>
-</g>
-<!-- Node43 -->
-<g id="node44" class="node">
-<title>Node43</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="868.5,-364 868.5,-383 909.5,-383 909.5,-364 868.5,-364"/>
-<text text-anchor="middle" x="889" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tuple</text>
+<path fill="none" stroke="#191970" d="M1697.8912,-426.6988C1650.3698,-414.641 1584.455,-397.3024 1542.2039,-385.7333"/>
+<polygon fill="#191970" stroke="#191970" points="1543.0905,-382.3472 1532.5202,-383.0678 1541.2327,-389.0962 1543.0905,-382.3472"/>
</g>
<!-- Node39->Node43 -->
-<g id="edge118" class="edge">
+<g id="edge116" class="edge">
<title>Node39->Node43</title>
-<path fill="none" stroke="#191970" d="M1023.2477,-425.4441C994.6603,-415.5933 956.9143,-402.1533 924,-389 922.3878,-388.3557 920.7402,-387.6815 919.0811,-386.9902"/>
-<polygon fill="#191970" stroke="#191970" points="920.1914,-383.6581 909.6214,-382.9331 917.4322,-390.0914 920.1914,-383.6581"/>
+<path fill="none" stroke="#191970" d="M1814.1085,-435.1435C1847.4355,-432.1015 1890.0626,-428.2613 1928,-425 2123.488,-408.1947 2172.9678,-410.4572 2368,-389 2394.951,-386.0349 2425.424,-381.7572 2447.7726,-378.4464"/>
+<polygon fill="#191970" stroke="#191970" points="2448.3616,-381.8974 2457.7331,-376.9553 2447.3251,-374.9745 2448.3616,-381.8974"/>
+</g>
+<!-- Node44 -->
+<g id="node45" class="node">
+<title>Node44</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1696.5,-364 1696.5,-383 1737.5,-383 1737.5,-364 1696.5,-364"/>
+<text text-anchor="middle" x="1717" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tuple</text>
+</g>
+<!-- Node39->Node44 -->
+<g id="edge119" class="edge">
+<title>Node39->Node44</title>
+<path fill="none" stroke="#191970" d="M1747.1503,-425.2967C1741.3689,-415.3645 1733.7858,-402.3371 1727.6709,-391.8321"/>
+<polygon fill="#191970" stroke="#191970" points="1730.6175,-389.9367 1722.5618,-383.055 1724.5677,-393.4582 1730.6175,-389.9367"/>
</g>
<!-- Node40->Node16 -->
<g id="edge96" class="edge">
<title>Node40->Node16</title>
-<path fill="none" stroke="#191970" d="M1554.0455,-370.4296C1763.7799,-360.1689 2425.3687,-327.4104 2435,-322 2466.4918,-304.3094 2470.2971,-289.4984 2481,-255 2483.6339,-246.5103 2482.0861,-243.8223 2481,-235 2474.7308,-184.0747 2477.9497,-167.8353 2453,-123 2445.5411,-109.5961 2433.6951,-97.1754 2423.4054,-87.8914"/>
-<polygon fill="#191970" stroke="#191970" points="2425.4589,-85.0405 2415.6005,-81.1594 2420.8869,-90.3412 2425.4589,-85.0405"/>
+<path fill="none" stroke="#191970" d="M893.7279,-359.8783C890.4407,-359.2292 887.1831,-358.5992 884,-358 748.3486,-332.4634 646.9698,-381.1777 591,-255 540.7259,-141.6626 763.3385,-91.4734 849.3263,-76.545"/>
+<polygon fill="#191970" stroke="#191970" points="850.0626,-79.9703 859.3429,-74.859 848.9007,-73.0674 850.0626,-79.9703"/>
</g>
<!-- Node40->Node20 -->
<g id="edge94" class="edge">
<title>Node40->Node20</title>
-<path fill="none" stroke="#191970" d="M1486.4111,-358.4896C1481.931,-340.6085 1477.67,-310.6878 1492,-291 1521.5272,-250.4331 1556.884,-282.2722 1599,-255 1618.5223,-242.3584 1635.9916,-222.0616 1647.1798,-207.3106"/>
-<polygon fill="#191970" stroke="#191970" points="1650.2105,-209.0961 1653.2966,-198.9607 1644.5635,-204.9594 1650.2105,-209.0961"/>
+<path fill="none" stroke="#191970" d="M1020.1488,-368.7058C1126.6286,-360.2207 1337.2896,-341.5968 1409,-322 1426.608,-317.1881 1543.3996,-264.4777 1559,-255 1582.9834,-240.4294 1607.8324,-219.811 1624.0133,-205.5126"/>
+<polygon fill="#191970" stroke="#191970" points="1626.5269,-207.9597 1631.6342,-198.6771 1621.8529,-202.7487 1626.5269,-207.9597"/>
</g>
<!-- Node40->Node22 -->
<g id="edge97" class="edge">
<title>Node40->Node22</title>
-<path fill="none" stroke="#191970" d="M1500.7856,-358.4659C1513.8194,-339.6957 1538.6737,-307.994 1568,-291 1594.61,-275.58 1677.2571,-259.6225 1726.323,-251.1535"/>
-<polygon fill="#191970" stroke="#191970" points="1726.9807,-254.5919 1736.2512,-249.4628 1725.8055,-247.6913 1726.9807,-254.5919"/>
+<path fill="none" stroke="#191970" d="M1020.0177,-360.7387C1024.7477,-359.8041 1029.4485,-358.8826 1034,-358 1119.6617,-341.3899 1147.8452,-358.7198 1227,-322 1246.772,-312.8278 1247.2258,-303.615 1265,-291 1280.7359,-279.8317 1299.2048,-268.4842 1313.8776,-259.8541"/>
+<polygon fill="#191970" stroke="#191970" points="1316.0196,-262.6573 1322.9046,-254.6044 1312.5006,-256.6061 1316.0196,-262.6573"/>
</g>
<!-- Node40->Node29 -->
<g id="edge98" class="edge">
<title>Node40->Node29</title>
-<path fill="none" stroke="#191970" d="M1531.8444,-358.4639C1558.6799,-348.5849 1593.9782,-335.5905 1622.5629,-325.0675"/>
-<polygon fill="#191970" stroke="#191970" points="1623.8023,-328.3409 1631.9775,-321.6017 1621.384,-321.7719 1623.8023,-328.3409"/>
+<path fill="none" stroke="#191970" d="M907.6277,-358.4639C874.625,-348.4131 831.0337,-335.1375 796.1727,-324.5208"/>
+<polygon fill="#191970" stroke="#191970" points="797.1735,-321.1669 786.5876,-321.6017 795.1342,-327.8633 797.1735,-321.1669"/>
</g>
<!-- Node40->Node34 -->
<g id="edge95" class="edge">
<title>Node40->Node34</title>
-<path fill="none" stroke="#191970" d="M1427.8437,-370.1161C1354.4829,-364.9158 1230.4244,-352.1726 1128,-322 1075.5446,-306.5475 1018.5238,-277.3034 986.5285,-259.64"/>
-<polygon fill="#191970" stroke="#191970" points="987.8719,-256.3813 977.4351,-254.5582 984.457,-262.4919 987.8719,-256.3813"/>
+<path fill="none" stroke="#191970" d="M894.8242,-358.4323C876.2377,-350.7135 857.9201,-339.1702 847,-322 839.6061,-310.3742 840.4587,-303.1259 847,-291 854.573,-276.9617 868.5073,-266.5084 881.9614,-259.1453"/>
+<polygon fill="#191970" stroke="#191970" points="883.758,-262.1593 891.115,-254.5353 880.6094,-255.9074 883.758,-262.1593"/>
</g>
<!-- Node41->Node8 -->
<g id="edge106" class="edge">
<title>Node41->Node8</title>
-<path fill="none" stroke="#191970" d="M1049.7659,-363.9096C1003.2307,-337.4525 880.2221,-263.5124 906,-235 969.0089,-165.3073 1627.8257,-140.9469 1846.1525,-134.7534"/>
-<polygon fill="#191970" stroke="#191970" points="1846.3137,-138.2504 1856.2119,-134.4721 1846.118,-131.2531 1846.3137,-138.2504"/>
+<path fill="none" stroke="#191970" d="M1564.1969,-369.0183C1663.1891,-361.3489 1845.5921,-344.3471 1865,-322 1874.0342,-311.5976 1865.4911,-304.769 1865,-291 1864.1105,-266.0636 1875.2102,-255.5107 1861,-235 1827.0232,-185.9584 1796.0717,-196.3877 1739,-179 1668.0281,-157.3774 1583.7655,-145.3398 1525.7476,-139.0558"/>
+<polygon fill="#191970" stroke="#191970" points="1525.9013,-135.5527 1515.5896,-137.9853 1525.1676,-142.5141 1525.9013,-135.5527"/>
</g>
<!-- Node41->Node9 -->
<g id="edge103" class="edge">
<title>Node41->Node9</title>
-<path fill="none" stroke="#191970" d="M1023.0637,-363.927C995.8657,-356.2322 961.5684,-343.1351 937,-322 910.8799,-299.53 906.9767,-288.2653 898,-255 889.193,-222.3633 908.7972,-211.8926 901,-179 893.6983,-148.1973 876.1132,-116.1855 862.9813,-95.1941"/>
-<polygon fill="#191970" stroke="#191970" points="865.838,-93.165 857.4918,-86.6387 859.9465,-96.9452 865.838,-93.165"/>
+<path fill="none" stroke="#191970" d="M1437.5949,-369.0877C1312.1828,-360.0507 1041.1412,-338.8668 1002,-322 982.6091,-313.6441 984.5974,-300.9982 966,-291 949.8044,-282.293 437.3509,-139.9874 254.2434,-89.2746"/>
+<polygon fill="#191970" stroke="#191970" points="255.0476,-85.8656 244.4762,-86.5699 253.1794,-92.6118 255.0476,-85.8656"/>
</g>
<!-- Node41->Node14 -->
-<g id="edge108" class="edge">
+<g id="edge109" class="edge">
<title>Node41->Node14</title>
-<path fill="none" stroke="#191970" d="M1022.5321,-363.9716C969.4153,-352.385 883.861,-333.0157 853,-322 809.8943,-306.6136 522.9774,-163.3886 482,-143 464.5599,-134.3225 461.223,-129.8832 443,-123 384.3309,-100.8396 312.5655,-85.1619 271.9855,-77.2914"/>
-<polygon fill="#191970" stroke="#191970" points="272.6044,-73.8464 262.1263,-75.4143 271.2952,-80.7229 272.6044,-73.8464"/>
+<path fill="none" stroke="#191970" d="M1564.2657,-368.1045C1600.2427,-365.0672 1646.1381,-361.2445 1687,-358 1898.9359,-341.172 1959.2148,-379.123 2164,-322 2211.1459,-308.8491 2513.3205,-145.0857 2557,-123 2581.9109,-110.4042 2610.2541,-96.0955 2630.7299,-85.7626"/>
+<polygon fill="#191970" stroke="#191970" points="2632.4703,-88.8048 2639.8213,-81.175 2629.3168,-82.5553 2632.4703,-88.8048"/>
</g>
<!-- Node41->Node18 -->
-<g id="edge110" class="edge">
+<g id="edge111" class="edge">
<title>Node41->Node18</title>
-<path fill="none" stroke="#191970" d="M1130.2382,-370.6541C1287.5902,-363.2951 1688.1733,-342.7578 1745,-322 1767.1296,-313.9165 1766.5165,-300.6707 1788,-291 1854.4653,-261.0809 1877.1885,-272.2777 1948,-255 1949.4497,-254.6463 1950.9292,-254.28 1952.4227,-253.9058"/>
-<polygon fill="#191970" stroke="#191970" points="1953.4971,-257.2438 1962.3168,-251.3732 1951.7612,-250.4624 1953.4971,-257.2438"/>
+<path fill="none" stroke="#191970" d="M1510.4595,-363.8504C1526.6494,-347.6345 1561.1148,-314.3995 1594,-291 1610.788,-279.0545 1630.9564,-267.7055 1647.0903,-259.2689"/>
+<polygon fill="#191970" stroke="#191970" points="1648.7728,-262.3394 1656.0673,-254.6556 1645.5733,-256.1134 1648.7728,-262.3394"/>
</g>
<!-- Node41->Node21 -->
<g id="edge107" class="edge">
<title>Node41->Node21</title>
-<path fill="none" stroke="#191970" d="M1066.098,-363.8416C1064.1435,-342.9114 1059.4829,-293.004 1056.8655,-264.9762"/>
-<polygon fill="#191970" stroke="#191970" points="1060.3271,-264.4017 1055.9124,-254.7705 1053.3575,-265.0526 1060.3271,-264.4017"/>
+<path fill="none" stroke="#191970" d="M1564.1563,-368.5251C1658.9217,-360.4121 1828.8076,-343.1007 1847,-322 1861.3493,-305.3568 1849.0857,-280.1817 1837.446,-263.1902"/>
+<polygon fill="#191970" stroke="#191970" points="1839.9537,-260.6914 1831.2166,-254.6985 1834.3096,-264.8319 1839.9537,-260.6914"/>
</g>
<!-- Node41->Node23 -->
<g id="edge105" class="edge">
<title>Node41->Node23</title>
-<path fill="none" stroke="#191970" d="M1013.0395,-363.9439C978.8717,-356.2331 934.7446,-343.1124 900,-322 864.2787,-300.294 851.8709,-292.7861 834,-255 827.2262,-240.6775 825.7095,-222.7004 825.7568,-209.1264"/>
-<polygon fill="#191970" stroke="#191970" points="829.2612,-209.0437 826.0834,-198.9367 822.2648,-208.8194 829.2612,-209.0437"/>
+<path fill="none" stroke="#191970" d="M1437.648,-370.8426C1328.574,-365.5385 1110.2544,-351.5189 1040,-322 1019.9058,-313.557 1021.4607,-300.8154 1002,-291 944.2524,-261.8739 900.6754,-306.0781 861,-255 835.2723,-221.8782 888.1372,-205.0767 938.4274,-196.7489"/>
+<polygon fill="#191970" stroke="#191970" points="939.0836,-200.189 948.427,-195.194 938.008,-193.2721 939.0836,-200.189"/>
</g>
<!-- Node41->Node31 -->
<g id="edge104" class="edge">
<title>Node41->Node31</title>
-<path fill="none" stroke="#191970" d="M1090.7838,-363.9005C1115.6446,-353.8663 1155.2953,-337.8627 1186.4215,-325.2997"/>
-<polygon fill="#191970" stroke="#191970" points="1187.8187,-328.5102 1195.7819,-321.5218 1185.1987,-322.019 1187.8187,-328.5102"/>
+<path fill="none" stroke="#191970" d="M1477.5028,-363.9005C1453.0473,-353.9096 1414.1059,-338.0006 1383.4154,-325.4624"/>
+<polygon fill="#191970" stroke="#191970" points="1384.3507,-322.0637 1373.7697,-321.5218 1381.7033,-328.5438 1384.3507,-322.0637"/>
</g>
<!-- Node41->Node34 -->
-<g id="edge109" class="edge">
+<g id="edge110" class="edge">
<title>Node41->Node34</title>
-<path fill="none" stroke="#191970" d="M1055.8181,-363.8215C1044.6367,-353.8798 1027.2447,-337.6834 1014,-322 998.1317,-303.2099 982.4775,-279.6827 972.3448,-263.6116"/>
-<polygon fill="#191970" stroke="#191970" points="975.0457,-261.3264 966.7967,-254.6777 969.0991,-265.0194 975.0457,-261.3264"/>
+<path fill="none" stroke="#191970" d="M1437.9947,-367.4815C1329.3605,-356.8461 1115.2927,-334.6595 1083,-322 1061.0655,-313.4012 1060.7736,-302.114 1040,-291 1014.2928,-277.2465 983.5761,-265.7551 959.2695,-257.6978"/>
+<polygon fill="#191970" stroke="#191970" points="960.2098,-254.3233 949.6177,-254.5661 958.0493,-260.9816 960.2098,-254.3233"/>
</g>
<!-- Node41->Node36 -->
<g id="edge102" class="edge">
<title>Node41->Node36</title>
-<path fill="none" stroke="#191970" d="M1130.2356,-370.6945C1369.4459,-360.0756 2205.56,-322.8994 2207,-322 2227.513,-309.188 2238.7394,-282.6345 2244.3064,-264.3223"/>
-<polygon fill="#191970" stroke="#191970" points="2247.6871,-265.2285 2246.9577,-254.6588 2240.9365,-263.3764 2247.6871,-265.2285"/>
+<path fill="none" stroke="#191970" d="M1437.6843,-369.1368C1307.111,-359.8743 1016.5503,-337.8023 974,-322 951.5748,-313.6718 952.1866,-299.9443 930,-291 884.0149,-272.4616 743.2194,-256.2467 672.4782,-249.012"/>
+<polygon fill="#191970" stroke="#191970" points="672.461,-245.4926 662.1599,-247.9697 671.7575,-252.4571 672.461,-245.4926"/>
</g>
<!-- Node41->Node39 -->
-<g id="edge111" class="edge">
+<g id="edge112" class="edge">
<title>Node41->Node39</title>
-<path fill="none" stroke="#191970" d="M1072.0346,-383.055C1073.7828,-391.4031 1074.529,-403.9072 1074.1918,-415.1199"/>
-<polygon fill="#191970" stroke="#191970" points="1070.6864,-415.0917 1073.5339,-425.2967 1077.6718,-415.5433 1070.6864,-415.0917"/>
+<path fill="none" stroke="#191970" d="M1542.0591,-383.0053C1583.3682,-393.3291 1647.3935,-410.0507 1694.6893,-422.806"/>
+<polygon fill="#191970" stroke="#191970" points="1693.9465,-426.2308 1704.5136,-425.4639 1695.7747,-419.4737 1693.9465,-426.2308"/>
</g>
-<!-- Node44->Node6 -->
-<g id="edge126" class="edge">
-<title>Node44->Node6</title>
-<path fill="none" stroke="#191970" d="M2605.5661,-492.2954C2622.4442,-476.9661 2650.0519,-446.6672 2633,-425 2606.5512,-391.3925 2334.8508,-340.0258 2207.8175,-317.6094"/>
-<polygon fill="#191970" stroke="#191970" points="2208.2778,-314.1367 2197.8228,-315.8528 2207.066,-321.0311 2208.2778,-314.1367"/>
+<!-- Node42 -->
+<g id="node43" class="node">
+<title>Node42</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="1494,-297 1494,-316 1542,-316 1542,-297 1494,-297"/>
+<text text-anchor="middle" x="1518" y="-304" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">mutex</text>
</g>
-<!-- Node45 -->
-<g id="node46" class="node">
-<title>Node45</title>
-<polygon fill="#ffffff" stroke="#bfbfbf" points="2564,-431 2564,-450 2624,-450 2624,-431 2564,-431"/>
-<text text-anchor="middle" x="2594" y="-438" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">iostream</text>
+<!-- Node41->Node42 -->
+<g id="edge108" class="edge">
+<title>Node41->Node42</title>
+<path fill="none" stroke="#191970" d="M1503.4357,-363.9005C1505.9352,-354.0495 1509.8945,-338.4451 1513.0545,-325.9912"/>
+<polygon fill="#191970" stroke="#191970" points="1516.5019,-326.6353 1515.5688,-316.0817 1509.7169,-324.9137 1516.5019,-326.6353"/>
</g>
-<!-- Node44->Node45 -->
+<!-- Node45->Node6 -->
<g id="edge127" class="edge">
-<title>Node44->Node45</title>
-<path fill="none" stroke="#191970" d="M2594,-492.3906C2594,-483.8657 2594,-471.1392 2594,-460.4235"/>
-<polygon fill="#191970" stroke="#191970" points="2597.5001,-460.2448 2594,-450.2449 2590.5001,-460.2449 2597.5001,-460.2448"/>
+<title>Node45->Node6</title>
+<path fill="none" stroke="#191970" d="M2314.9151,-492.3833C2311.5131,-466.7464 2298.7283,-396.1278 2258,-358 2232.8677,-334.4724 2196.7954,-321.686 2165.5637,-314.7405"/>
+<polygon fill="#191970" stroke="#191970" points="2166.115,-311.2805 2155.6136,-312.6845 2164.6985,-318.1357 2166.115,-311.2805"/>
</g>
-<!-- Node46->Node2 -->
-<g id="edge140" class="edge">
-<title>Node46->Node2</title>
-<path fill="none" stroke="#191970" d="M409.9811,-660.2455C399.7902,-651.9746 384.9638,-639.9416 372.9893,-630.2232"/>
-<polygon fill="#191970" stroke="#191970" points="374.8514,-627.2268 364.8812,-623.6427 370.4402,-632.662 374.8514,-627.2268"/>
+<!-- Node46 -->
+<g id="node47" class="node">
+<title>Node46</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="2426,-431 2426,-450 2486,-450 2486,-431 2426,-431"/>
+<text text-anchor="middle" x="2456" y="-438" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">iostream</text>
</g>
-<!-- Node46->Node3 -->
+<!-- Node45->Node46 -->
+<g id="edge128" class="edge">
+<title>Node45->Node46</title>
+<path fill="none" stroke="#191970" d="M2337.875,-492.3906C2361.2996,-482.1005 2398.6596,-465.6888 2425.0841,-454.0809"/>
+<polygon fill="#191970" stroke="#191970" points="2426.5307,-457.2683 2434.2786,-450.0419 2423.7154,-450.8594 2426.5307,-457.2683"/>
+</g>
+<!-- Node47->Node2 -->
<g id="edge141" class="edge">
-<title>Node46->Node3</title>
-<path fill="none" stroke="#191970" d="M425.3364,-660.2202C428.4062,-650.9695 432.9343,-636.6565 436,-624 439.6897,-608.7673 442.8164,-591.3106 444.9523,-578.1779"/>
-<polygon fill="#191970" stroke="#191970" points="448.4702,-578.3365 446.5635,-567.9147 441.5549,-577.2508 448.4702,-578.3365"/>
+<title>Node47->Node2</title>
+<path fill="none" stroke="#191970" d="M2588.616,-660.3733C2624.648,-650.9001 2679.7878,-636.4032 2718.8114,-626.1435"/>
+<polygon fill="#191970" stroke="#191970" points="2720.0804,-629.4289 2728.8618,-623.5011 2718.3005,-622.6589 2720.0804,-629.4289"/>
</g>
-<!-- Node46->Node8 -->
-<g id="edge144" class="edge">
-<title>Node46->Node8</title>
-<path fill="none" stroke="#191970" d="M462.2193,-668.1292C650.0998,-659.2764 1460.1177,-619.8506 2123,-568 2446.5527,-542.6917 2842,-765.041 2842,-440.5 2842,-440.5 2842,-440.5 2842,-373.5 2842,-262.4367 2738.526,-272.5416 2634,-235 2544.7064,-202.9293 2521.3872,-195.7542 2428,-179 2271.4666,-150.9171 2084.7789,-139.6419 1985.7583,-135.3823"/>
-<polygon fill="#191970" stroke="#191970" points="1985.7178,-131.8776 1975.5801,-134.956 1985.4248,-138.8715 1985.7178,-131.8776"/>
+<!-- Node47->Node3 -->
+<g id="edge142" class="edge">
+<title>Node47->Node3</title>
+<path fill="none" stroke="#191970" d="M2527.7863,-660.4509C2477.065,-640.4482 2359.5691,-594.1117 2301.5924,-571.2477"/>
+<polygon fill="#191970" stroke="#191970" points="2302.7392,-567.9377 2292.1524,-567.5249 2300.1711,-574.4496 2302.7392,-567.9377"/>
</g>
-<!-- Node46->Node14 -->
+<!-- Node47->Node8 -->
<g id="edge145" class="edge">
-<title>Node46->Node14</title>
-<path fill="none" stroke="#191970" d="M384.1297,-660.4412C287.2523,-633.4759 38,-547.2182 38,-373.5 38,-373.5 38,-373.5 38,-189 38,-111.2758 151.1239,-84.0783 208.0604,-75.3118"/>
-<polygon fill="#191970" stroke="#191970" points="208.5986,-78.7705 217.9979,-73.8813 207.6012,-71.8419 208.5986,-78.7705"/>
+<title>Node47->Node8</title>
+<path fill="none" stroke="#191970" d="M2511.8525,-669.2707C2192.515,-663.343 76,-621.6542 76,-558 76,-558 76,-558 76,-306.5 76,-173.0023 1105.7747,-140.6769 1386.2101,-134.349"/>
+<polygon fill="#191970" stroke="#191970" points="1386.3403,-137.847 1396.2602,-134.1261 1386.185,-130.8488 1386.3403,-137.847"/>
</g>
-<!-- Node46->Node19 -->
-<g id="edge142" class="edge">
-<title>Node46->Node19</title>
-<path fill="none" stroke="#191970" d="M462.0899,-669.5062C641.5091,-666.7775 1362,-650.0867 1362,-558 1362,-558 1362,-558 1362,-440.5 1362,-402.4372 1368.0512,-358.6682 1372.4668,-331.7965"/>
-<polygon fill="#191970" stroke="#191970" points="1375.9558,-332.1556 1374.1786,-321.7109 1369.0545,-330.9842 1375.9558,-332.1556"/>
+<!-- Node47->Node14 -->
+<g id="edge146" class="edge">
+<title>Node47->Node14</title>
+<path fill="none" stroke="#191970" d="M2567.4343,-660.3517C2631.3555,-619.0703 2872,-450.042 2872,-245 2872,-245 2872,-245 2872,-189 2872,-107.5262 2751.0403,-82.1202 2691.427,-74.5464"/>
+<polygon fill="#191970" stroke="#191970" points="2691.7324,-71.0583 2681.3922,-73.3668 2690.9151,-78.0104 2691.7324,-71.0583"/>
</g>
-<!-- Node46->Node26 -->
+<!-- Node47->Node19 -->
<g id="edge143" class="edge">
-<title>Node46->Node26</title>
-<path fill="none" stroke="#191970" d="M419.3989,-660.0885C415.9365,-647.1141 409.4921,-623.7053 403,-604 367.0244,-494.8041 247.36,-444.5827 323,-358 342.4657,-335.7182 518.1287,-319.5486 623.677,-311.7466"/>
-<polygon fill="#191970" stroke="#191970" points="623.9461,-315.2364 633.6649,-311.0178 623.4366,-308.2549 623.9461,-315.2364"/>
+<title>Node47->Node19</title>
+<path fill="none" stroke="#191970" d="M2511.79,-669.4098C2334.3314,-666.2841 1630,-648.1971 1630,-558 1630,-558 1630,-558 1630,-440.5 1630,-385.893 1686.1535,-347.357 1728.5712,-326.0372"/>
+<polygon fill="#191970" stroke="#191970" points="1730.1933,-329.1404 1737.6589,-321.6227 1727.1347,-322.844 1730.1933,-329.1404"/>
+</g>
+<!-- Node47->Node26 -->
+<g id="edge144" class="edge">
+<title>Node47->Node26</title>
+<path fill="none" stroke="#191970" d="M2552,-660.4116C2552,-641.0538 2552,-595.9167 2552,-558 2552,-558 2552,-558 2552,-502 2552,-435.4564 2558.4125,-404.6916 2511,-358 2493.6103,-340.8747 2436.8274,-326.9522 2389.4889,-317.921"/>
+<polygon fill="#191970" stroke="#191970" points="2390.0202,-314.4598 2379.5483,-316.0687 2388.7379,-321.3414 2390.0202,-314.4598"/>
</g>
</g>
</svg>
diff --git a/docs/reference/api/doxygen/algorithm_8h__incl.svg b/docs/reference/api/doxygen/algorithm_8h__incl.svg
index c5d5f556e..2d0f55fff 100644
--- a/docs/reference/api/doxygen/algorithm_8h__incl.svg
+++ b/docs/reference/api/doxygen/algorithm_8h__incl.svg
@@ -43,33 +43,33 @@
<path fill="none" stroke="#191970" d="M395.5546,-1019.4441C268.9561,-1011.6741 0,-985.6738 0,-905 0,-905 0,-905 0,-133 0,-96.1996 17.8333,-84.8767 50,-67 96.081,-41.3904 464.1678,-22.2273 575.9329,-16.9541"/>
<polygon fill="#191970" stroke="#191970" points="576.1035,-20.45 585.9293,-16.4874 575.777,-13.4577 576.1035,-20.45"/>
</g>
-<!-- Node49 -->
+<!-- Node50 -->
<g id="node35" class="node">
-<title>Node49</title>
+<title>Node50</title>
<g id="a_node35"><a xlink:href="relay_2base_8h.html" target="_top" xlink:title="Base classes for the Relay IR. ">
<polygon fill="#ffffff" stroke="#000000" points="690.5,-839.5 690.5,-858.5 789.5,-858.5 789.5,-839.5 690.5,-839.5"/>
<text text-anchor="middle" x="740" y="-846.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/relay/base.h</text>
</a>
</g>
</g>
-<!-- Node0->Node49 -->
+<!-- Node0->Node50 -->
<g id="edge129" class="edge">
-<title>Node0->Node49</title>
+<title>Node0->Node50</title>
<path fill="none" stroke="#191970" d="M495.8443,-1007.4832C538.6376,-988.7804 611.506,-954.4412 668,-915 688.8819,-900.4213 710.0433,-880.2324 723.9691,-866.0337"/>
<polygon fill="#191970" stroke="#191970" points="726.4893,-868.4624 730.9135,-858.8354 721.4515,-863.6022 726.4893,-868.4624"/>
</g>
-<!-- Node53 -->
+<!-- Node54 -->
<g id="node39" class="node">
-<title>Node53</title>
+<title>Node54</title>
<g id="a_node39"><a xlink:href="relay_2expr_8h.html" target="_top" xlink:title="Relay expression language. ">
<polygon fill="#ffffff" stroke="#000000" points="828.5,-951.5 828.5,-970.5 925.5,-970.5 925.5,-951.5 828.5,-951.5"/>
<text text-anchor="middle" x="877" y="-958.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/relay/expr.h</text>
</a>
</g>
</g>
-<!-- Node0->Node53 -->
+<!-- Node0->Node54 -->
<g id="edge160" class="edge">
-<title>Node0->Node53</title>
+<title>Node0->Node54</title>
<path fill="none" stroke="#191970" d="M524.2424,-1013.0254C604.4847,-1001.1911 740.6399,-981.1107 818.2345,-969.6669"/>
<polygon fill="#191970" stroke="#191970" points="818.9392,-973.1009 828.3215,-968.1792 817.9178,-966.1758 818.9392,-973.1009"/>
</g>
@@ -299,30 +299,30 @@
<path fill="none" stroke="#191970" d="M1500.7156,-611.1773C1574.068,-605.6001 1735.5809,-591.6187 1870,-568 2094.5483,-528.5447 2677.2856,-366.0907 2724,-344 2793.6187,-311.0781 2862.02,-250.244 2896.3005,-217.2359"/>
<polygon fill="#191970" stroke="#191970" points="2899.1491,-219.3472 2903.8704,-209.8624 2894.2648,-214.3328 2899.1491,-219.3472"/>
</g>
-<!-- Node44 -->
+<!-- Node45 -->
<g id="node30" class="node">
-<title>Node44</title>
+<title>Node45</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="826,-252 826,-271 870,-271 870,-252 826,-252"/>
<text text-anchor="middle" x="848" y="-259" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">limits</text>
</g>
-<!-- Node3->Node44 -->
+<!-- Node3->Node45 -->
<g id="edge117" class="edge">
-<title>Node3->Node44</title>
+<title>Node3->Node45</title>
<path fill="none" stroke="#191970" d="M1440.6597,-604.4643C1405.955,-587.8076 1333.6589,-551.4101 1278,-512 1236.3705,-482.5236 1238.0693,-457.8974 1192,-436 1103.3079,-393.8434 1060.2669,-446.9199 974,-400 920.6612,-370.9895 878.0803,-310.5811 858.9541,-280.0364"/>
<polygon fill="#191970" stroke="#191970" points="861.7844,-277.9556 853.5845,-271.2464 855.8107,-281.6047 861.7844,-277.9556"/>
</g>
-<!-- Node48 -->
+<!-- Node49 -->
<g id="node34" class="node">
-<title>Node48</title>
+<title>Node49</title>
<g id="a_node34"><a xlink:href="ir_2type_8h.html" target="_top" xlink:title="IR/AST nodes for the unified type system in TVM. ">
<polygon fill="#ffffff" stroke="#000000" points="2444,-548.5 2444,-567.5 2524,-567.5 2524,-548.5 2444,-548.5"/>
<text text-anchor="middle" x="2484" y="-555.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/type.h</text>
</a>
</g>
</g>
-<!-- Node3->Node48 -->
+<!-- Node3->Node49 -->
<g id="edge106" class="edge">
-<title>Node3->Node48</title>
+<title>Node3->Node49</title>
<path fill="none" stroke="#191970" d="M1500.7152,-605.9759C1505.5131,-605.2049 1510.357,-604.5185 1515,-604 1919.4669,-558.8299 2025.5331,-613.1701 2430,-568 2431.2792,-567.8571 2432.5736,-567.7015 2433.8782,-567.5349"/>
<polygon fill="#191970" stroke="#191970" points="2434.5956,-570.9677 2443.991,-566.0712 2433.5928,-564.0399 2434.5956,-570.9677"/>
</g>
@@ -432,18 +432,18 @@
<path fill="none" stroke="#191970" d="M1579.7895,-439.1886C1672.8308,-426.5804 1879.4589,-399.1475 2054,-380 2224.8327,-361.2593 2276.9504,-401.2272 2439,-344 2489.7934,-326.0625 2511.754,-322.8682 2540,-277 2563.8297,-238.3033 2554.3222,-181.9569 2546.806,-152.5365"/>
<polygon fill="#191970" stroke="#191970" points="2550.1202,-151.388 2544.0965,-142.6722 2543.3702,-153.2421 2550.1202,-151.388"/>
</g>
-<!-- Node46 -->
+<!-- Node47 -->
<g id="node32" class="node">
-<title>Node46</title>
+<title>Node47</title>
<g id="a_node32"><a xlink:href="repr__printer_8h.html" target="_top" xlink:title="Printer class to print repr string of each AST/IR nodes. ">
<polygon fill="#ffffff" stroke="#000000" points="1697.5,-380.5 1697.5,-399.5 1828.5,-399.5 1828.5,-380.5 1697.5,-380.5"/>
<text text-anchor="middle" x="1763" y="-387.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/repr_printer.h</text>
</a>
</g>
</g>
-<!-- Node5->Node46 -->
+<!-- Node5->Node47 -->
<g id="edge92" class="edge">
-<title>Node5->Node46</title>
+<title>Node5->Node47</title>
<path fill="none" stroke="#191970" d="M1569.7894,-436.4369C1609.498,-426.8932 1670.6104,-412.2052 1713.4455,-401.9101"/>
<polygon fill="#191970" stroke="#191970" points="1714.4695,-405.2637 1723.3747,-399.5237 1712.8337,-398.4575 1714.4695,-405.2637"/>
</g>
@@ -941,21 +941,21 @@
<path fill="none" stroke="#191970" d="M2272.9585,-313.4639C2310.1571,-301.5284 2363.6633,-285.0456 2401.5671,-273.8678"/>
<polygon fill="#191970" stroke="#191970" points="2402.7203,-277.1771 2411.332,-271.0053 2400.7512,-270.4597 2402.7203,-277.1771"/>
</g>
-<!-- Node41->Node44 -->
+<!-- Node41->Node45 -->
<g id="edge82" class="edge">
-<title>Node41->Node44</title>
+<title>Node41->Node45</title>
<path fill="none" stroke="#191970" d="M2171.939,-325.6852C1936.9807,-314.2943 1060.6383,-271.8088 880.4973,-263.0755"/>
<polygon fill="#191970" stroke="#191970" points="880.4981,-259.5715 870.3403,-262.5831 880.1591,-266.5633 880.4981,-259.5715"/>
</g>
-<!-- Node45 -->
+<!-- Node46 -->
<g id="node31" class="node">
-<title>Node45</title>
+<title>Node46</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="2172.5,-252 2172.5,-271 2213.5,-271 2213.5,-252 2172.5,-252"/>
<text text-anchor="middle" x="2193" y="-259" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tuple</text>
</g>
-<!-- Node41->Node45 -->
+<!-- Node41->Node46 -->
<g id="edge85" class="edge">
-<title>Node41->Node45</title>
+<title>Node41->Node46</title>
<path fill="none" stroke="#191970" d="M2221.6041,-313.2967C2216.1192,-303.3645 2208.9249,-290.3371 2203.1237,-279.8321"/>
<polygon fill="#191970" stroke="#191970" points="2206.1748,-278.1169 2198.2766,-271.055 2200.047,-281.5009 2206.1748,-278.1169"/>
</g>
@@ -1031,336 +1031,336 @@
<path fill="none" stroke="#191970" d="M2421.0965,-271.0053C2389.874,-281.2643 2336.5762,-297.841 2294.1424,-310.5654"/>
<polygon fill="#191970" stroke="#191970" points="2293.0193,-307.2479 2284.4396,-313.4639 2295.023,-313.955 2293.0193,-307.2479"/>
</g>
-<!-- Node46->Node8 -->
+<!-- Node47->Node8 -->
<g id="edge93" class="edge">
-<title>Node46->Node8</title>
+<title>Node47->Node8</title>
<path fill="none" stroke="#191970" d="M1753.1316,-380.3051C1721.3691,-349.1013 1621.5151,-251.0036 1581.1665,-211.3646"/>
<polygon fill="#191970" stroke="#191970" points="1583.5144,-208.7647 1573.928,-204.2534 1578.6087,-213.7582 1583.5144,-208.7647"/>
</g>
-<!-- Node47 -->
+<!-- Node48 -->
<g id="node33" class="node">
-<title>Node47</title>
+<title>Node48</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="2852,-319 2852,-338 2912,-338 2912,-319 2852,-319"/>
<text text-anchor="middle" x="2882" y="-326" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">iostream</text>
</g>
-<!-- Node46->Node47 -->
+<!-- Node47->Node48 -->
<g id="edge94" class="edge">
-<title>Node46->Node47</title>
+<title>Node47->Node48</title>
<path fill="none" stroke="#191970" d="M1828.6285,-386.4955C1970.9606,-378.882 2318.6813,-360.2111 2610,-344 2692.3499,-339.4175 2788.7136,-333.8862 2841.7923,-330.8249"/>
<polygon fill="#191970" stroke="#191970" points="2842.0322,-334.3169 2851.8139,-330.2465 2841.6288,-327.3286 2842.0322,-334.3169"/>
</g>
-<!-- Node48->Node4 -->
+<!-- Node49->Node4 -->
<g id="edge107" class="edge">
-<title>Node48->Node4</title>
+<title>Node49->Node4</title>
<path fill="none" stroke="#191970" d="M2443.9203,-555.6473C2291.7672,-546.7159 1749.9797,-514.9129 1580.8301,-504.9837"/>
<polygon fill="#191970" stroke="#191970" points="1580.8906,-501.4814 1570.7026,-504.3893 1580.4803,-508.4693 1580.8906,-501.4814"/>
</g>
-<!-- Node48->Node5 -->
+<!-- Node49->Node5 -->
<g id="edge108" class="edge">
-<title>Node48->Node5</title>
+<title>Node49->Node5</title>
<path fill="none" stroke="#191970" d="M2443.9429,-554.9311C2371.5336,-549.045 2214.1918,-534.7031 2083,-512 2044.3325,-505.3085 2035.7199,-498.3814 1997,-492 1851.8263,-468.0741 1679.0651,-455.0126 1589.8909,-449.3941"/>
<polygon fill="#191970" stroke="#191970" points="1589.8963,-445.8878 1579.6987,-448.7615 1589.4626,-452.8744 1589.8963,-445.8878"/>
</g>
-<!-- Node48->Node10 -->
+<!-- Node49->Node10 -->
<g id="edge111" class="edge">
-<title>Node48->Node10</title>
+<title>Node49->Node10</title>
<path fill="none" stroke="#191970" d="M2494.7212,-548.2786C2538.9508,-506.9742 2703.69,-340.7348 2677,-179 2672.7207,-153.0683 2678.2809,-139.7168 2658,-123 2614.38,-87.0457 2467.1999,-94.6257 2397.5861,-87.4301"/>
<polygon fill="#191970" stroke="#191970" points="2397.8539,-83.9369 2387.503,-86.1973 2397.0043,-90.8851 2397.8539,-83.9369"/>
</g>
-<!-- Node48->Node16 -->
+<!-- Node49->Node16 -->
<g id="edge112" class="edge">
-<title>Node48->Node16</title>
+<title>Node49->Node16</title>
<path fill="none" stroke="#191970" d="M2443.9298,-555.9245C2255.5787,-545.7207 1457.9197,-497.3968 1231,-400 1004.3302,-302.7105 1007.0698,-178.4169 787,-67 738.8161,-42.6055 677.3159,-28.1212 640.3826,-20.9993"/>
<polygon fill="#191970" stroke="#191970" points="640.6717,-17.493 630.1999,-19.1029 639.39,-24.3747 640.6717,-17.493"/>
</g>
-<!-- Node48->Node21 -->
+<!-- Node49->Node21 -->
<g id="edge109" class="edge">
-<title>Node48->Node21</title>
+<title>Node49->Node21</title>
<path fill="none" stroke="#191970" d="M2505.6374,-548.3446C2553.5779,-527.0741 2672.5302,-474.9703 2774,-436 2933.902,-374.5883 3125.8788,-310.3116 3218.9411,-279.7421"/>
<polygon fill="#191970" stroke="#191970" points="3220.3026,-282.9791 3228.7136,-276.5366 3218.1208,-276.3278 3220.3026,-282.9791"/>
</g>
-<!-- Node48->Node28 -->
+<!-- Node49->Node28 -->
<g id="edge110" class="edge">
-<title>Node48->Node28</title>
+<title>Node49->Node28</title>
<path fill="none" stroke="#191970" d="M2443.659,-551.7944C2350.4199,-536.3129 2114.4703,-490.5556 1938,-400 1851.7936,-355.7632 1816.5826,-354.1785 1758,-277 1743.7626,-258.2432 1736.4864,-231.7877 1732.9702,-213.9347"/>
<polygon fill="#191970" stroke="#191970" points="1736.4158,-213.3196 1731.2381,-204.0761 1729.5214,-214.531 1736.4158,-213.3196"/>
</g>
-<!-- Node49->Node4 -->
+<!-- Node50->Node4 -->
<g id="edge130" class="edge">
-<title>Node49->Node4</title>
+<title>Node50->Node4</title>
<path fill="none" stroke="#191970" d="M789.5913,-841.6545C827.1796,-834.8719 879.4464,-822.7911 922,-803 1020.7535,-757.071 1025.2064,-711.82 1121,-660 1248.0302,-591.2825 1411.4245,-537.6361 1487.5053,-514.4669"/>
<polygon fill="#191970" stroke="#191970" points="1488.6545,-517.776 1497.2137,-511.5318 1486.6287,-511.0755 1488.6545,-517.776"/>
</g>
-<!-- Node49->Node5 -->
+<!-- Node50->Node5 -->
<g id="edge131" class="edge">
-<title>Node49->Node5</title>
+<title>Node50->Node5</title>
<path fill="none" stroke="#191970" d="M753.5753,-839.2257C817.7432,-793.4438 1098.2403,-597.9915 1355,-492 1392.1319,-476.6718 1435.8335,-465.1792 1470.2763,-457.5149"/>
<polygon fill="#191970" stroke="#191970" points="1471.1634,-460.9039 1480.1902,-455.3571 1469.6746,-454.064 1471.1634,-460.9039"/>
</g>
-<!-- Node49->Node16 -->
+<!-- Node50->Node16 -->
<g id="edge158" class="edge">
-<title>Node49->Node16</title>
+<title>Node50->Node16</title>
<path fill="none" stroke="#191970" d="M690.1448,-839.7948C638.1523,-829.8714 560.3527,-814.0202 532,-803 484.2467,-784.4391 375.0305,-714.3303 337,-680 312.4511,-657.8396 307.6304,-650.6153 288,-624 253.2962,-576.9477 248.4143,-562.5279 219,-512 187.8777,-458.5381 152,-451.8609 152,-390 152,-390 152,-390 152,-133 152,-98.4675 163.9121,-85.6114 193,-67 256.4113,-26.4273 489.4212,-17.793 575.6407,-15.9768"/>
<polygon fill="#191970" stroke="#191970" points="575.9426,-19.4718 585.8731,-15.7793 575.8074,-12.4731 575.9426,-19.4718"/>
</g>
-<!-- Node49->Node20 -->
+<!-- Node50->Node20 -->
<g id="edge159" class="edge">
-<title>Node49->Node20</title>
+<title>Node50->Node20</title>
<path fill="none" stroke="#191970" d="M711.6481,-839.4273C633.3463,-812.5045 416.561,-734.5376 366,-680 344.7797,-657.1107 342,-645.2125 342,-614 342,-614 342,-614 342,-558 342,-440.3901 359.9946,-397.3043 442,-313 570.3651,-181.0363 1186.0047,-142.1724 1337.3416,-134.5395"/>
<polygon fill="#191970" stroke="#191970" points="1337.6067,-138.0308 1347.4224,-134.0431 1337.2623,-131.0393 1337.6067,-138.0308"/>
</g>
-<!-- Node50 -->
+<!-- Node51 -->
<g id="node36" class="node">
-<title>Node50</title>
+<title>Node51</title>
<g id="a_node36"><a xlink:href="tir_2expr_8h.html" target="_top" xlink:title="TIR expressions. ">
<polygon fill="#ffffff" stroke="#000000" points="2382.5,-778 2382.5,-797 2465.5,-797 2465.5,-778 2382.5,-778"/>
<text text-anchor="middle" x="2424" y="-785" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/tir/expr.h</text>
</a>
</g>
</g>
-<!-- Node49->Node50 -->
+<!-- Node50->Node51 -->
<g id="edge132" class="edge">
-<title>Node49->Node50</title>
+<title>Node50->Node51</title>
<path fill="none" stroke="#191970" d="M789.6811,-846.4856C832.6396,-844.3523 896.4087,-841.2835 952,-839 1508.051,-816.1593 2182.5281,-794.9332 2372.1962,-789.0843"/>
<polygon fill="#191970" stroke="#191970" points="2372.5421,-792.5754 2382.4295,-788.7691 2372.3265,-785.5787 2372.5421,-792.5754"/>
</g>
-<!-- Node50->Node3 -->
+<!-- Node51->Node3 -->
<g id="edge133" class="edge">
-<title>Node50->Node3</title>
+<title>Node51->Node3</title>
<path fill="none" stroke="#191970" d="M2382.3347,-779.0643C2370.5928,-776.7294 2357.8035,-774.2262 2346,-772 2028.3614,-712.0915 1645.4448,-645.7067 1510.7597,-622.5334"/>
<polygon fill="#191970" stroke="#191970" points="1511.1927,-619.0566 1500.7442,-620.8112 1510.0064,-625.9553 1511.1927,-619.0566"/>
</g>
-<!-- Node50->Node5 -->
+<!-- Node51->Node5 -->
<g id="edge135" class="edge">
-<title>Node50->Node5</title>
+<title>Node51->Node5</title>
<path fill="none" stroke="#191970" d="M2382.3427,-786.4723C2284.516,-783.5038 2032.7503,-772.6326 1826,-736 1638.7315,-702.8192 1528.5497,-775.0587 1413,-624 1368.9764,-566.4477 1460.882,-492.7205 1506.4309,-461.2928"/>
<polygon fill="#191970" stroke="#191970" points="1508.5443,-464.0889 1514.8686,-455.5886 1504.6238,-458.2897 1508.5443,-464.0889"/>
</g>
-<!-- Node50->Node8 -->
+<!-- Node51->Node8 -->
<g id="edge134" class="edge">
-<title>Node50->Node8</title>
+<title>Node51->Node8</title>
<path fill="none" stroke="#191970" d="M2382.1971,-787.1452C2180.5849,-785.2461 1313.71,-774.8658 1200,-736 1159.2754,-722.0804 1140.7622,-718.2322 1121,-680 1039.5367,-522.3995 1157.7446,-443.4332 1278,-313 1336.6152,-249.4239 1434.5214,-218.9543 1499.1781,-205.0895"/>
<polygon fill="#191970" stroke="#191970" points="1500.1748,-208.4575 1509.2585,-203.0043 1498.7568,-201.6026 1500.1748,-208.4575"/>
</g>
-<!-- Node50->Node11 -->
+<!-- Node51->Node11 -->
<g id="edge136" class="edge">
-<title>Node50->Node11</title>
+<title>Node51->Node11</title>
<path fill="none" stroke="#191970" d="M2465.5269,-784.28C2509.9291,-780.9091 2582.4106,-775.6132 2645,-772 2733.1661,-766.9103 3361.143,-777.6827 3439,-736 3488.5097,-709.4937 3738,-384.6586 3738,-328.5 3738,-328.5 3738,-328.5 3738,-261.5 3738,-69.8296 3040.2262,-26.7898 2806.9378,-17.7563"/>
<polygon fill="#191970" stroke="#191970" points="2806.9701,-14.2552 2796.8453,-17.3756 2806.7062,-21.2502 2806.9701,-14.2552"/>
</g>
-<!-- Node50->Node16 -->
+<!-- Node51->Node16 -->
<g id="edge155" class="edge">
-<title>Node50->Node16</title>
+<title>Node51->Node16</title>
<path fill="none" stroke="#191970" d="M2382.466,-786.5336C2181.3641,-781.7665 1307.4253,-760.0193 1036,-736 857.6257,-720.2151 808.3567,-731.9946 637,-680 423.7649,-615.2983 228,-612.8352 228,-390 228,-390 228,-390 228,-133 228,-101.7875 227.7859,-86.6952 252,-67 301.2779,-26.9184 497.0435,-18.0178 575.3647,-16.0526"/>
<polygon fill="#191970" stroke="#191970" points="575.6917,-19.5461 585.6092,-15.8188 575.5319,-12.5479 575.6917,-19.5461"/>
</g>
-<!-- Node50->Node18 -->
+<!-- Node51->Node18 -->
<g id="edge157" class="edge">
-<title>Node50->Node18</title>
+<title>Node51->Node18</title>
<path fill="none" stroke="#191970" d="M2465.5259,-784.2629C2509.9273,-780.8777 2582.4082,-775.57 2645,-772 2735.9401,-766.8131 3378.2146,-766.6246 3464,-736 3666.7566,-663.6178 3852,-661.2892 3852,-446 3852,-446 3852,-446 3852,-133 3852,-101.7875 3850.9503,-88.1544 3828,-67 3805.4447,-46.2097 3720.5546,-29.0384 3672.9271,-20.7865"/>
<polygon fill="#191970" stroke="#191970" points="3673.2478,-17.291 3662.8033,-19.0696 3672.0773,-24.1924 3673.2478,-17.291"/>
</g>
-<!-- Node50->Node21 -->
+<!-- Node51->Node21 -->
<g id="edge137" class="edge">
-<title>Node50->Node21</title>
+<title>Node51->Node21</title>
<path fill="none" stroke="#191970" d="M2465.5394,-784.4814C2509.952,-781.2787 2582.4421,-776.1219 2645,-772 2709.8714,-767.7257 3181.2207,-780.0875 3229,-736 3358.8142,-616.2165 3304.1553,-366.376 3282.3802,-286.5751"/>
<polygon fill="#191970" stroke="#191970" points="3285.7328,-285.5676 3279.6658,-276.8819 3278.9921,-287.4553 3285.7328,-285.5676"/>
</g>
-<!-- Node50->Node22 -->
+<!-- Node51->Node22 -->
<g id="edge152" class="edge">
-<title>Node50->Node22</title>
+<title>Node51->Node22</title>
<path fill="none" stroke="#191970" d="M2465.5354,-784.4203C2509.9447,-781.1665 2582.4321,-775.9675 2645,-772 2786.0544,-763.0555 3150.2624,-792.0759 3280,-736 3505.2469,-638.6426 3469.6735,-246.2434 3457.8095,-152.9664"/>
<polygon fill="#191970" stroke="#191970" points="3461.2584,-152.3498 3456.4694,-142.8991 3454.3196,-153.2736 3461.2584,-152.3498"/>
</g>
-<!-- Node50->Node28 -->
+<!-- Node51->Node28 -->
<g id="edge140" class="edge">
-<title>Node50->Node28</title>
+<title>Node51->Node28</title>
<path fill="none" stroke="#191970" d="M2435.026,-777.9897C2473.0937,-744.0296 2593.4307,-626.5846 2533,-548 2376.8562,-344.9493 2208.8722,-506.6818 1976,-400 1887.9097,-359.6447 1867.0444,-342.8873 1796,-277 1774.5837,-257.1383 1754.2734,-230.0641 1741.9798,-212.4355"/>
<polygon fill="#191970" stroke="#191970" points="1744.8556,-210.4404 1736.3184,-204.1661 1739.0795,-214.3949 1744.8556,-210.4404"/>
</g>
-<!-- Node50->Node33 -->
+<!-- Node51->Node33 -->
<g id="edge139" class="edge">
-<title>Node50->Node33</title>
+<title>Node51->Node33</title>
<path fill="none" stroke="#191970" d="M2465.8821,-784.7784C2571.5519,-777.6731 2846.9007,-757.6101 2936,-736 3001.93,-720.0094 3028.2959,-726.1733 3078,-680 3217.8135,-550.1184 3010.2933,-295.6081 2940.5233,-217.7092"/>
<polygon fill="#191970" stroke="#191970" points="2942.7733,-214.9792 2933.4688,-209.9119 2937.5825,-219.6756 2942.7733,-214.9792"/>
</g>
-<!-- Node50->Node36 -->
+<!-- Node51->Node36 -->
<g id="edge156" class="edge">
-<title>Node50->Node36</title>
+<title>Node51->Node36</title>
<path fill="none" stroke="#191970" d="M2465.5375,-784.4526C2509.9485,-781.2258 2582.4373,-776.0491 2645,-772 2712.4219,-767.6365 3193.063,-769.0331 3252,-736 3460.2821,-619.2617 3384.658,-243.1148 3362.9974,-152.6848"/>
<polygon fill="#191970" stroke="#191970" points="3366.3783,-151.7762 3360.5932,-142.9002 3359.5805,-153.4465 3366.3783,-151.7762"/>
</g>
-<!-- Node50->Node42 -->
+<!-- Node51->Node42 -->
<g id="edge138" class="edge">
-<title>Node50->Node42</title>
+<title>Node51->Node42</title>
<path fill="none" stroke="#191970" d="M2465.5278,-784.2945C2509.9307,-780.9357 2582.4128,-775.6498 2645,-772 2730.947,-766.988 3346.345,-782.1872 3419,-736 3434.8967,-725.8944 3567.532,-452.2845 3577,-436 3586.5433,-419.586 3594.6752,-418.2247 3600,-400 3610.844,-362.8851 3609.8068,-350.4024 3600,-313 3597.4772,-303.3784 3592.7069,-293.6592 3587.7528,-285.2958"/>
<polygon fill="#191970" stroke="#191970" points="3590.7007,-283.4089 3582.4026,-276.8216 3584.7817,-287.1459 3590.7007,-283.4089"/>
</g>
-<!-- Node50->Node44 -->
+<!-- Node51->Node45 -->
<g id="edge154" class="edge">
-<title>Node50->Node44</title>
+<title>Node51->Node45</title>
<path fill="none" stroke="#191970" d="M2382.4426,-786.9132C2188.847,-784.0179 1374.2138,-770.0043 1122,-736 1083.1097,-730.7567 1074.6316,-722.8958 1036,-716 936.1626,-698.179 647.6831,-706.7127 589,-624 492.8976,-488.5454 740.2251,-325.2905 822.4776,-276.1679"/>
<polygon fill="#191970" stroke="#191970" points="824.3909,-279.1029 831.2229,-271.0051 820.8322,-273.0749 824.3909,-279.1029"/>
</g>
-<!-- Node50->Node47 -->
+<!-- Node51->Node48 -->
<g id="edge153" class="edge">
-<title>Node50->Node47</title>
+<title>Node51->Node48</title>
<path fill="none" stroke="#191970" d="M2465.64,-785.5481C2594.4791,-777.0678 2978,-732.4684 2978,-502 2978,-502 2978,-502 2978,-446 2978,-401.0753 2935.3952,-363.8602 2906.9508,-344.0203"/>
<polygon fill="#191970" stroke="#191970" points="2908.5383,-340.8705 2898.2851,-338.2016 2904.636,-346.6819 2908.5383,-340.8705"/>
</g>
-<!-- Node51 -->
+<!-- Node52 -->
<g id="node37" class="node">
-<title>Node51</title>
+<title>Node52</title>
<g id="a_node37"><a xlink:href="buffer_8h.html" target="_top" xlink:title="Symbolic n-dimensional array, to represent a memory buffer. ">
<polygon fill="#ffffff" stroke="#000000" points="1834.5,-716.5 1834.5,-735.5 1923.5,-735.5 1923.5,-716.5 1834.5,-716.5"/>
<text text-anchor="middle" x="1879" y="-723.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/tir/buffer.h</text>
</a>
</g>
</g>
-<!-- Node50->Node51 -->
+<!-- Node51->Node52 -->
<g id="edge141" class="edge">
-<title>Node50->Node51</title>
+<title>Node51->Node52</title>
<path fill="none" stroke="#191970" d="M2382.1397,-782.7763C2284.7083,-771.7818 2042.7822,-744.4818 1933.7825,-732.1819"/>
<polygon fill="#191970" stroke="#191970" points="1933.9878,-728.6829 1923.6584,-731.0394 1933.2028,-735.6388 1933.9878,-728.6829"/>
</g>
-<!-- Node52 -->
+<!-- Node53 -->
<g id="node38" class="node">
-<title>Node52</title>
+<title>Node53</title>
<g id="a_node38"><a xlink:href="var_8h.html" target="_top" xlink:title="Variables in the TIR. ">
<polygon fill="#ffffff" stroke="#000000" points="1130.5,-660.5 1130.5,-679.5 1207.5,-679.5 1207.5,-660.5 1130.5,-660.5"/>
<text text-anchor="middle" x="1169" y="-667.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/tir/var.h</text>
</a>
</g>
</g>
-<!-- Node50->Node52 -->
+<!-- Node51->Node53 -->
<g id="edge151" class="edge">
-<title>Node50->Node52</title>
+<title>Node51->Node53</title>
<path fill="none" stroke="#191970" d="M2382.2511,-786.5805C2177.1891,-781.9525 1284.3544,-760.4623 1231,-736 1209.0191,-725.922 1191.07,-704.1286 1180.2428,-688.3358"/>
<polygon fill="#191970" stroke="#191970" points="1183.1107,-686.3261 1174.7143,-679.8645 1177.2486,-690.1518 1183.1107,-686.3261"/>
</g>
-<!-- Node51->Node3 -->
+<!-- Node52->Node3 -->
<g id="edge142" class="edge">
-<title>Node51->Node3</title>
+<title>Node52->Node3</title>
<path fill="none" stroke="#191970" d="M1843.3616,-716.4509C1767.4745,-696.1176 1590.0305,-648.5728 1506.2837,-626.1334"/>
<polygon fill="#191970" stroke="#191970" points="1507.1134,-622.7323 1496.5482,-623.5249 1505.3016,-629.4938 1507.1134,-622.7323"/>
</g>
-<!-- Node51->Node16 -->
+<!-- Node52->Node16 -->
<g id="edge150" class="edge">
-<title>Node51->Node16</title>
+<title>Node52->Node16</title>
<path fill="none" stroke="#191970" d="M1834.3571,-725.1101C1602.6243,-720.415 546.4213,-698.0551 518,-680 495.9582,-665.9976 304,-287.6134 304,-261.5 304,-261.5 304,-261.5 304,-133 304,-92.7538 330.9376,-86.7582 366,-67 401.6697,-46.8996 518.4113,-28.1772 575.7018,-19.9267"/>
<polygon fill="#191970" stroke="#191970" points="576.2661,-23.3818 585.6744,-18.5101 575.2815,-16.4513 576.2661,-23.3818"/>
</g>
-<!-- Node51->Node21 -->
+<!-- Node52->Node21 -->
<g id="edge143" class="edge">
-<title>Node51->Node21</title>
+<title>Node52->Node21</title>
<path fill="none" stroke="#191970" d="M1923.5845,-724.5747C2077.3995,-719.5108 2585.26,-701.5327 2747,-680 2885.3413,-661.5824 3054,-753.5619 3054,-614 3054,-614 3054,-614 3054,-558 3054,-498.0857 3182.7385,-392.4392 3218,-344 3232.264,-324.4054 3247.8712,-301.7239 3259.1422,-285.108"/>
<polygon fill="#191970" stroke="#191970" points="3262.2205,-286.8038 3264.9217,-276.5591 3256.4214,-282.8833 3262.2205,-286.8038"/>
</g>
-<!-- Node51->Node33 -->
+<!-- Node52->Node33 -->
<g id="edge144" class="edge">
-<title>Node51->Node33</title>
+<title>Node52->Node33</title>
<path fill="none" stroke="#191970" d="M1923.6617,-722.6965C2033.3545,-713.046 2322.3915,-678.2158 2533,-568 2705.3713,-477.7945 2729.9427,-421.6859 2860,-277 2876.5869,-258.5474 2892.7889,-235.4141 2904.0046,-218.343"/>
<polygon fill="#191970" stroke="#191970" points="2907.1994,-219.8482 2909.6933,-209.551 2901.3224,-216.0455 2907.1994,-219.8482"/>
</g>
-<!-- Node51->Node52 -->
+<!-- Node52->Node53 -->
<g id="edge145" class="edge">
-<title>Node51->Node52</title>
+<title>Node52->Node53</title>
<path fill="none" stroke="#191970" d="M1834.3174,-722.4757C1707.9933,-712.5121 1348.9375,-684.1923 1217.7562,-673.8456"/>
<polygon fill="#191970" stroke="#191970" points="1217.8533,-670.3424 1207.609,-673.0452 1217.3028,-677.3208 1217.8533,-670.3424"/>
</g>
-<!-- Node52->Node3 -->
+<!-- Node53->Node3 -->
<g id="edge146" class="edge">
-<title>Node52->Node3</title>
+<title>Node53->Node3</title>
<path fill="none" stroke="#191970" d="M1207.7753,-662.5636C1260.1683,-652.5157 1353.6188,-634.5936 1411.1668,-623.557"/>
<polygon fill="#191970" stroke="#191970" points="1411.9659,-626.9677 1421.1277,-621.6467 1410.6474,-620.093 1411.9659,-626.9677"/>
</g>
-<!-- Node52->Node5 -->
+<!-- Node53->Node5 -->
<g id="edge147" class="edge">
-<title>Node52->Node5</title>
+<title>Node53->Node5</title>
<path fill="none" stroke="#191970" d="M1178.3206,-660.4758C1207.3484,-631.3049 1299.7231,-542.1357 1392,-492 1419.3507,-477.1399 1452.3538,-465.9259 1479.2265,-458.2917"/>
<polygon fill="#191970" stroke="#191970" points="1480.4914,-461.5736 1489.201,-455.541 1478.6304,-454.8255 1480.4914,-461.5736"/>
</g>
-<!-- Node52->Node16 -->
+<!-- Node53->Node16 -->
<g id="edge149" class="edge">
-<title>Node52->Node16</title>
+<title>Node53->Node16</title>
<path fill="none" stroke="#191970" d="M1133.6651,-660.4024C1020.011,-627.6628 660.769,-509.3057 480,-277 456.5486,-246.8627 456,-232.6867 456,-194.5 456,-194.5 456,-194.5 456,-133 456,-72.8142 531.5179,-39.0962 576.0946,-24.4359"/>
<polygon fill="#191970" stroke="#191970" points="577.4082,-27.6918 585.8979,-21.3534 575.3085,-21.0142 577.4082,-27.6918"/>
</g>
-<!-- Node52->Node28 -->
+<!-- Node53->Node28 -->
<g id="edge148" class="edge">
-<title>Node52->Node28</title>
+<title>Node53->Node28</title>
<path fill="none" stroke="#191970" d="M1169.7744,-660.2549C1174.0049,-613.781 1198.4632,-412.798 1311,-313 1363.4792,-266.4612 1560.0427,-225.1262 1664.2257,-205.9048"/>
<polygon fill="#191970" stroke="#191970" points="1665.1105,-209.3011 1674.3174,-204.0588 1663.8508,-202.4154 1665.1105,-209.3011"/>
</g>
-<!-- Node53->Node1 -->
+<!-- Node54->Node1 -->
<g id="edge161" class="edge">
-<title>Node53->Node1</title>
+<title>Node54->Node1</title>
<path fill="none" stroke="#191970" d="M828.4858,-959.4967C708.2966,-954.6311 402.1007,-934.8833 342,-859 230.398,-718.091 607.7967,-681.2076 747.8011,-672.48"/>
<polygon fill="#191970" stroke="#191970" points="748.0823,-675.9695 757.8546,-671.8764 747.6628,-668.982 748.0823,-675.9695"/>
</g>
-<!-- Node53->Node3 -->
+<!-- Node54->Node3 -->
<g id="edge162" class="edge">
-<title>Node53->Node3</title>
+<title>Node54->Node3</title>
<path fill="none" stroke="#191970" d="M925.5306,-956.5357C963.6172,-951.2365 1016.7472,-939.7491 1057,-915 1157.0818,-853.4655 1136.4066,-785.6775 1231,-716 1289.8963,-672.617 1370.1587,-642.3448 1418.6313,-626.5948"/>
<polygon fill="#191970" stroke="#191970" points="1419.7176,-629.9221 1428.1796,-623.5468 1417.5888,-623.2536 1419.7176,-629.9221"/>
</g>
-<!-- Node53->Node16 -->
+<!-- Node54->Node16 -->
<g id="edge215" class="edge">
-<title>Node53->Node16</title>
+<title>Node54->Node16</title>
<path fill="none" stroke="#191970" d="M828.2654,-959.6617C649.2022,-954.1081 38,-929.2724 38,-849 38,-849 38,-849 38,-133 38,-98.4675 49.6688,-85.2256 79,-67 121.2391,-40.7538 467.4511,-22.1919 575.7477,-16.9838"/>
<polygon fill="#191970" stroke="#191970" points="575.9325,-20.4791 585.755,-16.5081 575.6001,-13.487 575.9325,-20.4791"/>
</g>
-<!-- Node53->Node18 -->
+<!-- Node54->Node18 -->
<g id="edge216" class="edge">
-<title>Node53->Node18</title>
+<title>Node54->Node18</title>
<path fill="none" stroke="#191970" d="M925.5208,-960.5543C1318.5079,-956.7911 3966,-928.4894 3966,-849 3966,-849 3966,-849 3966,-133 3966,-95.6561 3945.9408,-86.3482 3914,-67 3873.7885,-42.6418 3736.8273,-25.596 3673.1054,-18.791"/>
<polygon fill="#191970" stroke="#191970" points="3673.1247,-15.274 3662.8145,-17.7134 3672.3956,-22.2359 3673.1247,-15.274"/>
</g>
-<!-- Node53->Node40 -->
+<!-- Node54->Node40 -->
<g id="edge213" class="edge">
-<title>Node53->Node40</title>
+<title>Node54->Node40</title>
<path fill="none" stroke="#191970" d="M828.3049,-959.8175C665.7509,-954.5695 152,-926.9523 152,-787.5 152,-787.5 152,-787.5 152,-670 152,-501.0618 623.9184,-302.8607 783,-246 874.6616,-213.2373 1170.0386,-199.8313 1281.1783,-195.8632"/>
<polygon fill="#191970" stroke="#191970" points="1281.4738,-199.3552 1291.3456,-195.5084 1281.2296,-192.3594 1281.4738,-199.3552"/>
</g>
-<!-- Node53->Node49 -->
+<!-- Node54->Node50 -->
<g id="edge217" class="edge">
-<title>Node53->Node49</title>
+<title>Node54->Node50</title>
<path fill="none" stroke="#191970" d="M865.3195,-951.4509C841.7617,-932.1921 788.3437,-888.5218 759.4411,-864.8935"/>
<polygon fill="#191970" stroke="#191970" points="761.6084,-862.1445 751.651,-858.5249 757.1778,-867.564 761.6084,-862.1445"/>
</g>
-<!-- Node54 -->
+<!-- Node55 -->
<g id="node40" class="node">
-<title>Node54</title>
+<title>Node55</title>
<g id="a_node40"><a xlink:href="ir_2module_8h.html" target="_top" xlink:title="IRModule that holds the functions and type definitions. ">
<polygon fill="#ffffff" stroke="#000000" points="2653.5,-778 2653.5,-797 2748.5,-797 2748.5,-778 2653.5,-778"/>
<text text-anchor="middle" x="2701" y="-785" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/module.h</text>
</a>
</g>
</g>
-<!-- Node53->Node54 -->
+<!-- Node54->Node55 -->
<g id="edge163" class="edge">
-<title>Node53->Node54</title>
+<title>Node54->Node55</title>
<path fill="none" stroke="#191970" d="M925.7876,-956.3593C1181.5483,-932.0312 2366.4167,-819.3258 2643.111,-793.0064"/>
<polygon fill="#191970" stroke="#191970" points="2643.7067,-796.4657 2653.3303,-792.0344 2643.0438,-789.4971 2643.7067,-796.4657"/>
</g>
-<!-- Node62 -->
+<!-- Node63 -->
<g id="node45" class="node">
-<title>Node62</title>
+<title>Node63</title>
<g id="a_node45"><a xlink:href="ir_2op_8h.html" target="_top" xlink:title="Primitive operators(builtin intrinsics) and registry for them. ">
<polygon fill="#ffffff" stroke="#ff0000" points="589,-895.5 589,-914.5 659,-914.5 659,-895.5 589,-895.5"/>
<text text-anchor="middle" x="624" y="-902.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/op.h</text>
</a>
</g>
</g>
-<!-- Node53->Node62 -->
+<!-- Node54->Node63 -->
<g id="edge196" class="edge">
-<title>Node53->Node62</title>
+<title>Node54->Node63</title>
<path fill="none" stroke="#191970" d="M833.7952,-951.4369C787.9071,-941.2798 715.6945,-925.296 669.0976,-914.9821"/>
<polygon fill="#191970" stroke="#191970" points="669.7948,-911.5518 659.2747,-912.8078 668.282,-918.3863 669.7948,-911.5518"/>
</g>
-<!-- Node69 -->
+<!-- Node70 -->
<g id="node48" class="node">
-<title>Node69</title>
+<title>Node70</title>
<g id="a_node48"><a xlink:href="virtual__device_8h.html" target="_top" xlink:title="A compile time representation for where data is to be stored at runtime, and how to compile code to c...">
<polygon fill="#ffffff" stroke="#ff0000" points="1204.5,-772.5 1204.5,-802.5 1307.5,-802.5 1307.5,-772.5 1204.5,-772.5"/>
<text text-anchor="start" x="1212.5" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/target/virtual</text>
@@ -1368,402 +1368,402 @@
</a>
</g>
</g>
-<!-- Node53->Node69 -->
+<!-- Node54->Node70 -->
<g id="edge209" class="edge">
-<title>Node53->Node69</title>
+<title>Node54->Node70</title>
<path fill="none" stroke="#191970" d="M879.8226,-951.4561C887.2953,-927.9984 909.9595,-867.3747 952,-839 990.5881,-812.9554 1117.0828,-798.5276 1194.2124,-791.9542"/>
<polygon fill="#191970" stroke="#191970" points="1194.6148,-795.4329 1204.2903,-791.1159 1194.0345,-788.457 1194.6148,-795.4329"/>
</g>
-<!-- Node76 -->
+<!-- Node77 -->
<g id="node49" class="node">
-<title>Node76</title>
+<title>Node77</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="934.5,-895.5 934.5,-914.5 977.5,-914.5 977.5,-895.5 934.5,-895.5"/>
<text text-anchor="middle" x="956" y="-902.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stack</text>
</g>
-<!-- Node53->Node76 -->
+<!-- Node54->Node77 -->
<g id="edge214" class="edge">
-<title>Node53->Node76</title>
+<title>Node54->Node77</title>
<path fill="none" stroke="#191970" d="M890.7607,-951.2455C902.6552,-942.814 920.065,-930.4729 933.9085,-920.6598"/>
<polygon fill="#191970" stroke="#191970" points="936.2628,-923.2811 942.3969,-914.6427 932.2146,-917.5704 936.2628,-923.2811"/>
</g>
-<!-- Node77 -->
+<!-- Node78 -->
<g id="node50" class="node">
-<title>Node77</title>
+<title>Node78</title>
<g id="a_node50"><a xlink:href="relay_2type_8h.html" target="_top" xlink:title="Relay typed AST nodes. ">
<polygon fill="#ffffff" stroke="#ff0000" points="995.5,-895.5 995.5,-914.5 1048.5,-914.5 1048.5,-895.5 995.5,-895.5"/>
<text text-anchor="middle" x="1022" y="-902.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">./type.h</text>
</a>
</g>
</g>
-<!-- Node53->Node77 -->
+<!-- Node54->Node78 -->
<g id="edge218" class="edge">
-<title>Node53->Node77</title>
+<title>Node54->Node78</title>
<path fill="none" stroke="#191970" d="M901.9264,-951.3733C925.6202,-942.2226 961.4511,-928.3844 987.8068,-918.2057"/>
<polygon fill="#191970" stroke="#191970" points="989.3314,-921.3689 997.3989,-914.5011 986.8094,-914.8389 989.3314,-921.3689"/>
</g>
-<!-- Node54->Node3 -->
+<!-- Node55->Node3 -->
<g id="edge172" class="edge">
-<title>Node54->Node3</title>
+<title>Node55->Node3</title>
<path fill="none" stroke="#191970" d="M2653.3108,-780.7553C2518.4475,-761.6902 2122.7637,-705.815 1794,-660 1693.4551,-645.9885 1575.8054,-629.7805 1510.7148,-620.8298"/>
<polygon fill="#191970" stroke="#191970" points="1511.1137,-617.3518 1500.7301,-619.457 1510.1602,-624.2865 1511.1137,-617.3518"/>
</g>
-<!-- Node54->Node16 -->
+<!-- Node55->Node16 -->
<g id="edge191" class="edge">
-<title>Node54->Node16</title>
+<title>Node55->Node16</title>
<path fill="none" stroke="#191970" d="M2653.409,-783.8009C2607.6914,-780.3504 2536.6593,-775.259 2475,-772 2044.6369,-749.2532 1935.694,-769.0579 1506,-736 1429.7074,-730.1305 1411.2748,-722.0965 1335,-716 1242.4313,-708.6012 579.0703,-728.699 500,-680 409.2761,-624.1236 266,-368.0505 266,-261.5 266,-261.5 266,-261.5 266,-133 266,-93.0568 292.0043,-86.255 327,-67 369.6032,-43.5592 511.8005,-25.8133 576.0285,-18.7906"/>
<polygon fill="#191970" stroke="#191970" points="576.4313,-22.2675 585.9997,-17.7183 575.6828,-15.3076 576.4313,-22.2675"/>
</g>
-<!-- Node54->Node18 -->
+<!-- Node55->Node18 -->
<g id="edge194" class="edge">
-<title>Node54->Node18</title>
+<title>Node55->Node18</title>
<path fill="none" stroke="#191970" d="M2748.5684,-786.3809C2941.9344,-781.6697 3663.4266,-762.464 3762,-736 3823.8172,-719.404 3890,-734.0062 3890,-670 3890,-670 3890,-670 3890,-133 3890,-101.7875 3889.4031,-87.6523 3866,-67 3837.5118,-41.8604 3728.0884,-25.7875 3672.6281,-19.0838"/>
<polygon fill="#191970" stroke="#191970" points="3672.9543,-15.5981 3662.6136,-17.9044 3672.1356,-22.55 3672.9543,-15.5981"/>
</g>
-<!-- Node54->Node20 -->
+<!-- Node55->Node20 -->
<g id="edge195" class="edge">
-<title>Node54->Node20</title>
+<title>Node55->Node20</title>
<path fill="none" stroke="#191970" d="M2653.3945,-784.0536C2607.6664,-780.7861 2536.627,-775.8224 2475,-772 2437.3745,-769.6663 1149.9739,-704.1172 1121,-680 956.5666,-543.1292 1226.6678,-195.2395 1244,-179 1269.7958,-154.8304 1309.2471,-143.1752 1337.3281,-137.6768"/>
<polygon fill="#191970" stroke="#191970" points="1338.2555,-141.0664 1347.4815,-135.8577 1337.0209,-134.1762 1338.2555,-141.0664"/>
</g>
-<!-- Node54->Node21 -->
+<!-- Node55->Node21 -->
<g id="edge188" class="edge">
-<title>Node54->Node21</title>
+<title>Node55->Node21</title>
<path fill="none" stroke="#191970" d="M2748.6673,-787.1204C2920.1596,-785.423 3499.392,-776.7021 3571,-736 3644.8194,-694.0409 3641.6096,-624.6134 3605,-548 3542.0244,-416.2102 3387.1418,-320.6127 3314.1978,-281.3039"/>
<polygon fill="#191970" stroke="#191970" points="3315.7739,-278.1779 3305.3016,-276.5707 3312.4859,-284.3577 3315.7739,-278.1779"/>
</g>
-<!-- Node54->Node33 -->
+<!-- Node55->Node33 -->
<g id="edge190" class="edge">
-<title>Node54->Node33</title>
+<title>Node55->Node33</title>
<path fill="none" stroke="#191970" d="M2748.5187,-785.9502C2907.9825,-780.5218 3415.389,-761.3161 3439,-736 3545.4415,-621.8718 3536.5908,-536.8281 3104,-246 3094.1638,-239.3872 3038.1718,-224.2302 2990.3364,-212.0617"/>
<polygon fill="#191970" stroke="#191970" points="2991.1368,-208.654 2980.5837,-209.593 2989.4191,-215.4399 2991.1368,-208.654"/>
</g>
-<!-- Node54->Node36 -->
+<!-- Node55->Node36 -->
<g id="edge192" class="edge">
-<title>Node54->Node36</title>
+<title>Node55->Node36</title>
<path fill="none" stroke="#191970" d="M2748.784,-786.1866C2922.862,-781.2062 3517.1462,-762.4202 3598,-736 3649.3252,-719.2287 3700,-723.9959 3700,-670 3700,-670 3700,-670 3700,-390 3700,-287.5172 3580.3337,-339.3502 3499,-277 3447.4115,-237.4524 3396.4414,-179.4669 3372.3614,-150.6324"/>
<polygon fill="#191970" stroke="#191970" points="3374.8499,-148.1495 3365.7806,-142.6723 3369.4548,-152.6097 3374.8499,-148.1495"/>
</g>
-<!-- Node54->Node42 -->
+<!-- Node55->Node42 -->
<g id="edge189" class="edge">
-<title>Node54->Node42</title>
+<title>Node55->Node42</title>
<path fill="none" stroke="#191970" d="M2748.7161,-786.5497C2936.889,-782.6015 3621.7667,-766.2669 3714,-736 3764.5971,-719.3962 3814,-723.2518 3814,-670 3814,-670 3814,-670 3814,-390 3814,-312.2876 3715.6775,-281.4469 3644.1178,-269.2919"/>
<polygon fill="#191970" stroke="#191970" points="3644.6401,-265.831 3634.211,-267.6978 3643.528,-272.7421 3644.6401,-265.831"/>
</g>
-<!-- Node54->Node48 -->
+<!-- Node55->Node49 -->
<g id="edge181" class="edge">
-<title>Node54->Node48</title>
+<title>Node55->Node49</title>
<path fill="none" stroke="#191970" d="M2684.8152,-777.9656C2648.2067,-755.6194 2558.3216,-696.2505 2508,-624 2498.2128,-609.9478 2491.9408,-591.4482 2488.2727,-577.6112"/>
<polygon fill="#191970" stroke="#191970" points="2491.6254,-576.5843 2485.872,-567.6878 2484.8217,-578.2304 2491.6254,-576.5843"/>
</g>
-<!-- Node55 -->
+<!-- Node56 -->
<g id="node41" class="node">
-<title>Node55</title>
+<title>Node56</title>
<g id="a_node41"><a xlink:href="ir_2adt_8h.html" target="_top" xlink:title="Algebraic data type definitions. ">
<polygon fill="#ffffff" stroke="#ff0000" points="2664,-660.5 2664,-679.5 2738,-679.5 2738,-660.5 2664,-660.5"/>
<text text-anchor="middle" x="2701" y="-667.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/adt.h</text>
</a>
</g>
</g>
-<!-- Node54->Node55 -->
+<!-- Node55->Node56 -->
<g id="edge164" class="edge">
-<title>Node54->Node55</title>
+<title>Node55->Node56</title>
<path fill="none" stroke="#191970" d="M2701,-777.8845C2701,-758.6913 2701,-715.286 2701,-689.7663"/>
<polygon fill="#191970" stroke="#191970" points="2704.5001,-689.7483 2701,-679.7484 2697.5001,-689.7484 2704.5001,-689.7483"/>
</g>
-<!-- Node57 -->
+<!-- Node58 -->
<g id="node42" class="node">
-<title>Node57</title>
+<title>Node58</title>
<g id="a_node42"><a xlink:href="ir_2function_8h.html" target="_top" xlink:title="Function nodes. ">
<polygon fill="#ffffff" stroke="#000000" points="2829,-716.5 2829,-735.5 2927,-735.5 2927,-716.5 2829,-716.5"/>
<text text-anchor="middle" x="2878" y="-723.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/function.h</text>
</a>
</g>
</g>
-<!-- Node54->Node57 -->
+<!-- Node55->Node58 -->
<g id="edge173" class="edge">
-<title>Node54->Node57</title>
+<title>Node55->Node58</title>
<path fill="none" stroke="#191970" d="M2728.6563,-777.8906C2758.9211,-767.3749 2807.5855,-750.466 2841.0911,-738.8243"/>
<polygon fill="#191970" stroke="#191970" points="2842.2407,-742.1302 2850.538,-735.5419 2839.9431,-735.5179 2842.2407,-742.1302"/>
</g>
-<!-- Node58 -->
+<!-- Node59 -->
<g id="node43" class="node">
-<title>Node58</title>
+<title>Node59</title>
<g id="a_node43"><a xlink:href="source__map_8h.html" target="_top" xlink:title="A map from source names to source code. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1630.5,-548.5 1630.5,-567.5 1771.5,-567.5 1771.5,-548.5 1630.5,-548.5"/>
<text text-anchor="middle" x="1701" y="-555.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/parser/source_map.h</text>
</a>
</g>
</g>
-<!-- Node54->Node58 -->
+<!-- Node55->Node59 -->
<g id="edge182" class="edge">
-<title>Node54->Node58</title>
+<title>Node55->Node59</title>
<path fill="none" stroke="#191970" d="M2659.3044,-777.9309C2499.6677,-741.2942 1927.5263,-609.9878 1752.3438,-569.7834"/>
<polygon fill="#191970" stroke="#191970" points="1753.1258,-566.3719 1742.5963,-567.5463 1751.5599,-573.1946 1753.1258,-566.3719"/>
</g>
-<!-- Node61 -->
+<!-- Node62 -->
<g id="node44" class="node">
-<title>Node61</title>
+<title>Node62</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="1240,-716.5 1240,-735.5 1326,-735.5 1326,-716.5 1240,-716.5"/>
<text text-anchor="middle" x="1283" y="-723.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">unordered_set</text>
</g>
-<!-- Node54->Node61 -->
+<!-- Node55->Node62 -->
<g id="edge193" class="edge">
-<title>Node54->Node61</title>
+<title>Node55->Node62</title>
<path fill="none" stroke="#191970" d="M2653.4121,-783.741C2607.6967,-780.2472 2536.6662,-775.1255 2475,-772 1987.8074,-747.307 1864.6626,-769.5553 1378,-736 1364.5521,-735.0728 1350.0669,-733.7262 1336.5587,-732.3202"/>
<polygon fill="#191970" stroke="#191970" points="1336.7682,-728.8227 1326.4527,-731.2393 1336.0237,-735.783 1336.7682,-728.8227"/>
</g>
-<!-- Node55->Node3 -->
+<!-- Node56->Node3 -->
<g id="edge165" class="edge">
-<title>Node55->Node3</title>
+<title>Node56->Node3</title>
<path fill="none" stroke="#191970" d="M2663.9598,-668.1209C2623.2679,-666.0692 2555.9816,-662.7141 2498,-660 2118.2816,-642.2255 1660.2651,-622.5115 1510.7856,-616.1219"/>
<polygon fill="#191970" stroke="#191970" points="1510.7019,-612.6152 1500.5616,-615.6851 1510.403,-619.6089 1510.7019,-612.6152"/>
</g>
-<!-- Node55->Node5 -->
+<!-- Node56->Node5 -->
<g id="edge167" class="edge">
-<title>Node55->Node5</title>
+<title>Node56->Node5</title>
<path fill="none" stroke="#191970" d="M2691.5185,-660.2057C2667.5706,-636.1766 2601.6498,-574.3518 2533,-548 2358.7082,-481.0965 1778.4176,-454.8667 1589.7933,-447.9808"/>
<polygon fill="#191970" stroke="#191970" points="1589.9042,-444.4827 1579.7847,-447.6203 1589.6521,-451.4781 1589.9042,-444.4827"/>
</g>
-<!-- Node55->Node10 -->
+<!-- Node56->Node10 -->
<g id="edge170" class="edge">
-<title>Node55->Node10</title>
+<title>Node56->Node10</title>
<path fill="none" stroke="#191970" d="M2738.2565,-662.3873C2762.4163,-655.7866 2793.1588,-644.0393 2814,-624 2890.7147,-550.237 3048.1685,-268.7657 2991,-179 2959.3846,-129.3576 2927.8682,-138.1629 2871,-123 2783.6408,-99.7073 2524.0585,-85.5545 2398.1067,-79.8774"/>
<polygon fill="#191970" stroke="#191970" points="2397.9537,-76.3672 2387.8079,-79.4187 2397.6422,-83.3603 2397.9537,-76.3672"/>
</g>
-<!-- Node55->Node16 -->
+<!-- Node56->Node16 -->
<g id="edge171" class="edge">
-<title>Node55->Node16</title>
+<title>Node56->Node16</title>
<path fill="none" stroke="#191970" d="M2663.9228,-669.7886C2479.9311,-668.5701 1663.8073,-661.1303 1413,-624 1235.0275,-597.6524 1193.175,-573.2226 1024,-512 871.1024,-456.668 793.2388,-476.508 699,-344 627.7433,-243.8068 612.2095,-89.8839 608.8854,-35.4276"/>
<polygon fill="#191970" stroke="#191970" points="612.3627,-34.9063 608.3337,-25.1074 605.3726,-35.28 612.3627,-34.9063"/>
</g>
-<!-- Node55->Node21 -->
+<!-- Node56->Node21 -->
<g id="edge168" class="edge">
-<title>Node55->Node21</title>
+<title>Node56->Node21</title>
<path fill="none" stroke="#191970" d="M2738.1669,-665.666C2775.3713,-660.1389 2833.3492,-648.2269 2878,-624 2922.2164,-600.0088 3173.5301,-359.3233 3251.758,-283.9465"/>
<polygon fill="#191970" stroke="#191970" points="3254.4205,-286.2413 3259.19,-276.7808 3249.5618,-281.2021 3254.4205,-286.2413"/>
</g>
-<!-- Node55->Node33 -->
+<!-- Node56->Node33 -->
<g id="edge169" class="edge">
-<title>Node55->Node33</title>
+<title>Node56->Node33</title>
<path fill="none" stroke="#191970" d="M2726.9001,-660.3737C2763.5134,-644.8905 2826,-610.7402 2826,-558 2826,-558 2826,-558 2826,-446 2826,-414.7875 2829.2983,-403.3594 2850,-380 2873.4658,-353.5215 2900.0952,-372.5437 2921,-344 2947.5944,-307.6877 2937.2444,-251.9783 2927.7966,-219.808"/>
<polygon fill="#191970" stroke="#191970" points="2931.0106,-218.3594 2924.6818,-209.8626 2924.3306,-220.4516 2931.0106,-218.3594"/>
</g>
-<!-- Node55->Node48 -->
+<!-- Node56->Node49 -->
<g id="edge166" class="edge">
-<title>Node55->Node48</title>
+<title>Node56->Node49</title>
<path fill="none" stroke="#191970" d="M2682.4987,-660.4509C2644.3039,-640.7375 2556.5527,-595.4465 2511.6178,-572.2543"/>
<polygon fill="#191970" stroke="#191970" points="2512.946,-569.0012 2502.4545,-567.5249 2509.7354,-575.2215 2512.946,-569.0012"/>
</g>
-<!-- Node57->Node1 -->
+<!-- Node58->Node1 -->
<g id="edge174" class="edge">
-<title>Node57->Node1</title>
+<title>Node58->Node1</title>
<path fill="none" stroke="#191970" d="M2828.7383,-724.7965C2621.2244,-719.7084 1797.7516,-699.3349 1121,-680 1024.4489,-677.2415 911.6751,-673.6705 848.1674,-671.6274"/>
<polygon fill="#191970" stroke="#191970" points="848.1379,-668.1247 838.0303,-671.3008 847.9124,-675.1211 848.1379,-668.1247"/>
</g>
-<!-- Node57->Node3 -->
+<!-- Node58->Node3 -->
<g id="edge175" class="edge">
-<title>Node57->Node3</title>
+<title>Node58->Node3</title>
<path fill="none" stroke="#191970" d="M2828.9275,-717.3184C2745.5802,-702.9284 2570.8986,-674.306 2422,-660 2074.4727,-626.6101 1653.8015,-617.0791 1511.1041,-614.7055"/>
<polygon fill="#191970" stroke="#191970" points="1510.9743,-611.203 1500.9188,-614.5404 1510.8608,-618.2021 1510.9743,-611.203"/>
</g>
-<!-- Node57->Node16 -->
+<!-- Node58->Node16 -->
<g id="edge179" class="edge">
-<title>Node57->Node16</title>
+<title>Node58->Node16</title>
<path fill="none" stroke="#191970" d="M2828.6614,-725.3405C2519.1384,-721.1525 854.7738,-697.8661 749,-680 669.3119,-666.54 636.9024,-677.1249 576,-624 417.8781,-486.0711 380,-404.3259 380,-194.5 380,-194.5 380,-194.5 380,-133 380,-46.0732 512.5702,-23.3944 575.5337,-17.5253"/>
<polygon fill="#191970" stroke="#191970" points="576.0527,-20.994 585.7233,-16.6663 575.4646,-14.0188 576.0527,-20.994"/>
</g>
-<!-- Node57->Node17 -->
+<!-- Node58->Node17 -->
<g id="edge180" class="edge">
-<title>Node57->Node17</title>
+<title>Node58->Node17</title>
<path fill="none" stroke="#191970" d="M2875.5211,-716.2414C2869.2161,-693.3485 2850.4289,-635.7367 2814,-604 2490.5902,-322.2466 2245.6875,-538.8994 1906,-277 1816.726,-208.1697 1755.2051,-82.4023 1734.0761,-34.6135"/>
<polygon fill="#191970" stroke="#191970" points="1737.1697,-32.9498 1729.9807,-25.1671 1730.7473,-35.7342 1737.1697,-32.9498"/>
</g>
-<!-- Node57->Node21 -->
+<!-- Node58->Node21 -->
<g id="edge176" class="edge">
-<title>Node57->Node21</title>
+<title>Node58->Node21</title>
<path fill="none" stroke="#191970" d="M2927.0208,-720.9983C2985.2887,-714.2273 3078.1491,-700.5751 3106,-680 3134.4726,-658.9656 3247.2099,-433.7151 3258,-400 3270.1902,-361.9101 3273.7278,-315.3503 3274.703,-287.0844"/>
<polygon fill="#191970" stroke="#191970" points="3278.2077,-286.9457 3274.9702,-276.8577 3271.2101,-286.7629 3278.2077,-286.9457"/>
</g>
-<!-- Node57->Node33 -->
+<!-- Node58->Node33 -->
<g id="edge178" class="edge">
-<title>Node57->Node33</title>
+<title>Node58->Node33</title>
<path fill="none" stroke="#191970" d="M2927.0959,-721.7181C2990.276,-713.2074 3092,-688.0126 3092,-614 3092,-614 3092,-614 3092,-558 3092,-488.6439 3125.8046,-301.3412 3084,-246 3077.7911,-237.7807 3030.7773,-223.5633 2988.5159,-212.1463"/>
<polygon fill="#191970" stroke="#191970" points="2989.3129,-208.7365 2978.7481,-209.5335 2987.504,-215.4988 2989.3129,-208.7365"/>
</g>
-<!-- Node57->Node42 -->
+<!-- Node58->Node42 -->
<g id="edge177" class="edge">
-<title>Node57->Node42</title>
+<title>Node58->Node42</title>
<path fill="none" stroke="#191970" d="M2927.24,-720.2559C2989.5243,-712.4173 3092.7905,-697.3455 3127,-680 3327.2317,-578.4754 3500.2366,-358.2207 3553.8617,-285.4052"/>
<polygon fill="#191970" stroke="#191970" points="3557.0078,-287.0319 3560.08,-276.8922 3551.3552,-282.9029 3557.0078,-287.0319"/>
</g>
-<!-- Node58->Node4 -->
+<!-- Node59->Node4 -->
<g id="edge183" class="edge">
-<title>Node58->Node4</title>
+<title>Node59->Node4</title>
<path fill="none" stroke="#191970" d="M1671.6041,-548.3733C1643.1693,-539.0613 1599.9123,-524.8953 1568.6919,-514.671"/>
<polygon fill="#191970" stroke="#191970" points="1569.605,-511.2872 1559.0124,-511.5011 1567.4264,-517.9396 1569.605,-511.2872"/>
</g>
-<!-- Node58->Node16 -->
+<!-- Node59->Node16 -->
<g id="edge185" class="edge">
-<title>Node58->Node16</title>
+<title>Node59->Node16</title>
<path fill="none" stroke="#191970" d="M1651.9095,-548.4807C1607.2198,-539.6381 1539.4869,-525.7935 1481,-512 1385.7298,-489.5316 1362.6638,-480.8998 1268,-456 1234.2212,-447.115 1226.1402,-443.3755 1192,-436 1098.584,-415.8189 1067.7519,-437.8607 980,-400 896.6693,-364.0468 876.8399,-345.2337 817,-277 746.267,-196.3452 773.697,-143.0152 698,-67 681.3415,-50.2714 658.1325,-37.1673 639.4558,-28.3642"/>
<polygon fill="#191970" stroke="#191970" points="640.7144,-25.092 630.1609,-24.1563 637.8275,-31.469 640.7144,-25.092"/>
</g>
-<!-- Node58->Node18 -->
+<!-- Node59->Node18 -->
<g id="edge186" class="edge">
-<title>Node58->Node18</title>
+<title>Node59->Node18</title>
<path fill="none" stroke="#191970" d="M1771.6369,-549.5734C1994.8548,-522.7121 2704.9014,-435.4869 3290,-344 3373.8529,-330.8886 3599.9055,-330.2421 3666,-277 3696.8843,-252.1213 3700,-234.1584 3700,-194.5 3700,-194.5 3700,-194.5 3700,-133 3700,-93.7716 3673.1458,-54.4419 3655.3687,-32.7039"/>
<polygon fill="#191970" stroke="#191970" points="3657.9896,-30.3829 3648.8532,-25.0184 3652.6501,-34.9096 3657.9896,-30.3829"/>
</g>
-<!-- Node58->Node20 -->
+<!-- Node59->Node20 -->
<g id="edge187" class="edge">
-<title>Node58->Node20</title>
+<title>Node59->Node20</title>
<path fill="none" stroke="#191970" d="M1685.4039,-548.3187C1662.783,-534.5515 1619.2732,-509.095 1580,-492 1445.5031,-433.4557 1367.721,-465.1366 1285,-344 1243.3114,-282.9512 1221.0691,-239.1816 1264,-179 1280.9535,-155.2342 1312.8166,-143.6439 1337.3803,-138.055"/>
<polygon fill="#191970" stroke="#191970" points="1338.2987,-141.4397 1347.3959,-136.0091 1336.8977,-134.5813 1338.2987,-141.4397"/>
</g>
-<!-- Node58->Node41 -->
+<!-- Node59->Node41 -->
<g id="edge184" class="edge">
-<title>Node58->Node41</title>
+<title>Node59->Node41</title>
<path fill="none" stroke="#191970" d="M1723.057,-548.4309C1802.8733,-513.8036 2077.6249,-394.606 2185.7671,-347.6899"/>
<polygon fill="#191970" stroke="#191970" points="2187.3126,-350.8346 2195.0935,-343.6438 2184.5266,-344.4129 2187.3126,-350.8346"/>
</g>
-<!-- Node62->Node1 -->
+<!-- Node63->Node1 -->
<g id="edge198" class="edge">
-<title>Node62->Node1</title>
+<title>Node63->Node1</title>
<path fill="none" stroke="#191970" d="M618.2941,-895.2089C611.3718,-881.9639 601.6073,-857.8297 610,-839 643.1249,-764.6813 726.8723,-709.3706 770.5469,-684.5147"/>
<polygon fill="#191970" stroke="#191970" points="772.445,-687.4638 779.4777,-679.5397 769.0385,-681.3486 772.445,-687.4638"/>
</g>
-<!-- Node62->Node3 -->
+<!-- Node63->Node3 -->
<g id="edge199" class="edge">
-<title>Node62->Node3</title>
+<title>Node63->Node3</title>
<path fill="none" stroke="#191970" d="M629.7471,-895.4138C638.8422,-881.0531 657.9112,-854.0392 681,-839 837.9155,-736.7906 904.0372,-767.794 1084,-716 1100.2248,-711.3304 1319.4661,-652.1688 1415.955,-626.1459"/>
<polygon fill="#191970" stroke="#191970" points="1416.9797,-629.4947 1425.7233,-623.5115 1415.1569,-622.7362 1416.9797,-629.4947"/>
</g>
-<!-- Node62->Node16 -->
+<!-- Node63->Node16 -->
<g id="edge206" class="edge">
-<title>Node62->Node16</title>
+<title>Node63->Node16</title>
<path fill="none" stroke="#191970" d="M588.8395,-902.9751C518.3665,-898.398 362.6719,-885.525 317,-859 114.3444,-741.3029 76,-624.3542 76,-390 76,-390 76,-390 76,-133 76,-101.7875 75.3716,-86.1746 100,-67 137.5738,-37.7466 469.9536,-21.319 575.7852,-16.7956"/>
<polygon fill="#191970" stroke="#191970" points="576.0223,-20.2888 585.8663,-16.3716 575.7281,-13.295 576.0223,-20.2888"/>
</g>
-<!-- Node62->Node18 -->
+<!-- Node63->Node18 -->
<g id="edge207" class="edge">
-<title>Node62->Node18</title>
+<title>Node63->Node18</title>
<path fill="none" stroke="#191970" d="M659.1361,-903.7787C921.7735,-894.6141 2581.7647,-836.0676 3084,-803 3462.9196,-778.0516 3928,-1049.74 3928,-670 3928,-670 3928,-670 3928,-133 3928,-98.6999 3916.0086,-86.7993 3888,-67 3853.6175,-42.695 3731.8978,-25.9343 3672.7468,-19.0315"/>
<polygon fill="#191970" stroke="#191970" points="3673.134,-15.5531 3662.8014,-17.8953 3672.3394,-22.5078 3673.134,-15.5531"/>
</g>
-<!-- Node62->Node20 -->
+<!-- Node63->Node20 -->
<g id="edge208" class="edge">
-<title>Node62->Node20</title>
+<title>Node63->Node20</title>
<path fill="none" stroke="#191970" d="M588.9997,-903.504C522.1415,-899.9827 379.563,-889.0714 342,-859 271.3218,-802.4179 266,-760.5369 266,-670 266,-670 266,-670 266,-614 266,-422.5275 322.5043,-338.7735 490,-246 574.2203,-199.3516 601.479,-197.3005 696,-179 820.9664,-154.8048 1218.1484,-138.6068 1337.0506,-134.207"/>
<polygon fill="#191970" stroke="#191970" points="1337.4909,-137.6933 1347.3561,-133.8295 1337.2346,-130.698 1337.4909,-137.6933"/>
</g>
-<!-- Node62->Node48 -->
+<!-- Node63->Node49 -->
<g id="edge200" class="edge">
-<title>Node62->Node48</title>
+<title>Node63->Node49</title>
<path fill="none" stroke="#191970" d="M659.1676,-896.3677C694.4376,-887.5544 750.2252,-873.1898 798,-859 992.2793,-801.2962 1033.2569,-760.4052 1231,-716 1754.8866,-598.3559 1899.336,-649.811 2430,-568 2431.2721,-567.8039 2432.5602,-567.6006 2433.8591,-567.3916"/>
<polygon fill="#191970" stroke="#191970" points="2434.6601,-570.8061 2443.9412,-565.6961 2433.4992,-563.903 2434.6601,-570.8061"/>
</g>
-<!-- Node63 -->
+<!-- Node64 -->
<g id="node46" class="node">
-<title>Node63</title>
+<title>Node64</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="351.5,-839.5 351.5,-858.5 440.5,-858.5 440.5,-839.5 351.5,-839.5"/>
<text text-anchor="middle" x="396" y="-846.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dmlc/registry.h</text>
</g>
-<!-- Node62->Node63 -->
+<!-- Node63->Node64 -->
<g id="edge197" class="edge">
-<title>Node62->Node63</title>
+<title>Node63->Node64</title>
<path fill="none" stroke="#191970" d="M588.8747,-896.3727C550.2844,-886.8944 488.0424,-871.6069 444.741,-860.9715"/>
<polygon fill="#191970" stroke="#191970" points="445.5333,-857.5621 434.9871,-858.5758 443.8636,-864.3601 445.5333,-857.5621"/>
</g>
-<!-- Node64 -->
+<!-- Node65 -->
<g id="node47" class="node">
-<title>Node64</title>
+<title>Node65</title>
<g id="a_node47"><a xlink:href="type__relation_8h.html" target="_top" xlink:title="Type relation and function for type inference(checking). ">
<polygon fill="#ffffff" stroke="#ff0000" points="961,-839.5 961,-858.5 1083,-858.5 1083,-839.5 961,-839.5"/>
<text text-anchor="middle" x="1022" y="-846.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/type_relation.h</text>
</a>
</g>
</g>
-<!-- Node62->Node64 -->
+<!-- Node63->Node65 -->
<g id="edge201" class="edge">
-<title>Node62->Node64</title>
+<title>Node63->Node65</title>
<path fill="none" stroke="#191970" d="M659.3198,-900.0304C724.5592,-890.851 865.1033,-871.0759 950.8125,-859.0163"/>
<polygon fill="#191970" stroke="#191970" points="951.3124,-862.4806 960.7272,-857.6213 950.3371,-855.5488 951.3124,-862.4806"/>
</g>
-<!-- Node64->Node1 -->
+<!-- Node65->Node1 -->
<g id="edge202" class="edge">
-<title>Node64->Node1</title>
+<title>Node65->Node1</title>
<path fill="none" stroke="#191970" d="M1009.9568,-839.3762C973.4179,-810.1777 863.5478,-722.3797 818.0588,-686.0291"/>
<polygon fill="#191970" stroke="#191970" points="820.0998,-683.1799 810.1028,-679.6714 815.7299,-688.6484 820.0998,-683.1799"/>
</g>
-<!-- Node64->Node15 -->
+<!-- Node65->Node15 -->
<g id="edge205" class="edge">
-<title>Node64->Node15</title>
+<title>Node65->Node15</title>
<path fill="none" stroke="#191970" d="M1083.3365,-840.2585C1087.9539,-839.7731 1092.547,-839.3441 1097,-839 1833.2014,-782.1181 2020.9274,-847.1811 2758,-803 2909.7915,-793.9014 3336.8741,-848.6666 3439,-736 3495.9955,-673.1219 3915.7979,-900.1823 3138,-548 2810.1745,-399.5626 2688.0968,-454.408 2375,-277 2246.4697,-204.1717 2118.4675,-78.0725 2074.23,-32.5273"/>
<polygon fill="#191970" stroke="#191970" points="2076.6471,-29.9917 2067.1857,-25.2237 2071.6087,-34.8512 2076.6471,-29.9917"/>
</g>
-<!-- Node64->Node48 -->
+<!-- Node65->Node49 -->
<g id="edge204" class="edge">
-<title>Node64->Node48</title>
+<title>Node65->Node49</title>
<path fill="none" stroke="#191970" d="M1083.3884,-840.7642C1087.9938,-840.1619 1092.57,-839.5682 1097,-839 1468.3229,-791.3718 1567.4573,-816.7938 1933,-736 2133.8785,-691.601 2365.7052,-604.4373 2450.6295,-571.2544"/>
<polygon fill="#191970" stroke="#191970" points="2451.9837,-574.4828 2460.015,-567.5727 2449.4274,-567.9663 2451.9837,-574.4828"/>
</g>
-<!-- Node64->Node54 -->
+<!-- Node65->Node55 -->
<g id="edge203" class="edge">
-<title>Node64->Node54</title>
+<title>Node65->Node55</title>
<path fill="none" stroke="#191970" d="M1083.3382,-840.2809C1087.9553,-839.7903 1092.5478,-839.354 1097,-839 1707.7264,-790.446 1863.0317,-831.9664 2475,-803 2532.3397,-800.2859 2597.7624,-795.628 2643.4108,-792.1251"/>
<polygon fill="#191970" stroke="#191970" points="2643.7162,-795.612 2653.4166,-791.3515 2643.1765,-788.6329 2643.7162,-795.612"/>
</g>
-<!-- Node69->Node16 -->
+<!-- Node70->Node16 -->
<g id="edge210" class="edge">
-<title>Node69->Node16</title>
+<title>Node70->Node16</title>
<path fill="none" stroke="#191970" d="M1204.3128,-786.3573C1036.8213,-781.8757 511.4538,-761.4909 366,-680 300.234,-643.1544 190,-465.3841 190,-390 190,-390 190,-390 190,-133 190,-101.7875 189.6517,-86.529 214,-67 242.1182,-44.4474 486.9422,-24.3926 575.8834,-17.7894"/>
<polygon fill="#191970" stroke="#191970" points="576.2153,-21.2746 585.9321,-17.0516 575.7026,-14.2934 576.2153,-21.2746"/>
</g>
-<!-- Node69->Node18 -->
+<!-- Node70->Node18 -->
<g id="edge212" class="edge">
-<title>Node69->Node18</title>
+<title>Node70->Node18</title>
<path fill="none" stroke="#191970" d="M1307.7184,-786.5847C1572.6464,-781.8063 2771.6468,-759.0617 2936,-736 3036.3435,-721.92 3064.5202,-721.4071 3157,-680 3322.6881,-605.8146 3814,-376.0379 3814,-194.5 3814,-194.5 3814,-194.5 3814,-133 3814,-65.4788 3722.8974,-34.1521 3672.5335,-21.9928"/>
<polygon fill="#191970" stroke="#191970" points="3673.277,-18.5725 3662.7475,-19.747 3671.7112,-25.3951 3673.277,-18.5725"/>
</g>
-<!-- Node69->Node61 -->
+<!-- Node70->Node62 -->
<g id="edge211" class="edge">
-<title>Node69->Node61</title>
+<title>Node70->Node62</title>
<path fill="none" stroke="#191970" d="M1262.6742,-772.2977C1266.2659,-764.1166 1270.7324,-753.943 1274.5629,-745.2179"/>
<polygon fill="#191970" stroke="#191970" points="1277.8618,-746.4104 1278.677,-735.8469 1271.4522,-743.5964 1277.8618,-746.4104"/>
</g>
-<!-- Node77->Node1 -->
+<!-- Node78->Node1 -->
<g id="edge219" class="edge">
-<title>Node77->Node1</title>
+<title>Node78->Node1</title>
<path fill="none" stroke="#191970" d="M1004.2632,-895.3885C989.4422,-886.8959 968.1927,-873.6151 952,-859 890.853,-803.8105 833.6775,-723.3085 809.7717,-687.8724"/>
<polygon fill="#191970" stroke="#191970" points="812.6458,-685.8739 804.1808,-679.5024 806.8249,-689.762 812.6458,-685.8739"/>
</g>
-<!-- Node77->Node16 -->
+<!-- Node78->Node16 -->
<g id="edge223" class="edge">
-<title>Node77->Node16</title>
+<title>Node78->Node16</title>
<path fill="none" stroke="#191970" d="M995.3727,-897.113C992.2301,-896.3281 989.0551,-895.6009 986,-895 757.4298,-850.0408 679.5195,-904.8198 470,-803 252.0382,-697.0775 114,-632.3364 114,-390 114,-390 114,-390 114,-133 114,-95.9295 132.7254,-85.2366 165,-67 235.4454,-27.1952 486.1436,-18.0584 575.815,-16.0452"/>
<polygon fill="#191970" stroke="#191970" points="576.0049,-19.5421 585.9294,-15.8334 575.8583,-12.5436 576.0049,-19.5421"/>
</g>
-<!-- Node77->Node48 -->
+<!-- Node78->Node49 -->
<g id="edge220" class="edge">
-<title>Node77->Node48</title>
+<title>Node78->Node49</title>
<path fill="none" stroke="#191970" d="M1048.7414,-904.566C1240.7571,-901.2804 2411.7408,-878.3615 2758,-803 2840.5962,-785.0234 2887.895,-805.5068 2936,-736 2941.0586,-728.6909 2941.8725,-722.6728 2936,-716 2894.4087,-668.7404 2713.7156,-702.711 2655,-680 2591.5765,-655.468 2529.8223,-601.9459 2500.9553,-574.6818"/>
<polygon fill="#191970" stroke="#191970" points="2503.3512,-572.1301 2493.7114,-567.7341 2498.5058,-577.1821 2503.3512,-572.1301"/>
</g>
-<!-- Node77->Node49 -->
+<!-- Node78->Node50 -->
<g id="edge224" class="edge">
-<title>Node77->Node49</title>
+<title>Node78->Node50</title>
<path fill="none" stroke="#191970" d="M995.3292,-897.316C992.1944,-896.4947 989.0331,-895.7035 986,-895 922.8426,-880.3518 849.7317,-867.1512 799.7294,-858.6875"/>
<polygon fill="#191970" stroke="#191970" points="800.1767,-855.2136 789.7347,-857.0067 799.0158,-862.1167 800.1767,-855.2136"/>
</g>
-<!-- Node77->Node50 -->
+<!-- Node78->Node51 -->
<g id="edge222" class="edge">
-<title>Node77->Node50</title>
+<title>Node78->Node51</title>
<path fill="none" stroke="#191970" d="M1048.7069,-904.1275C1186.1046,-899.3919 1826.9678,-874.3044 2346,-803 2354.646,-801.8122 2363.7778,-800.266 2372.6284,-798.6108"/>
<polygon fill="#191970" stroke="#191970" points="2373.3073,-802.0444 2382.4606,-796.7089 2371.9779,-795.1718 2373.3073,-802.0444"/>
</g>
-<!-- Node77->Node64 -->
+<!-- Node78->Node65 -->
<g id="edge221" class="edge">
-<title>Node77->Node64</title>
+<title>Node78->Node65</title>
<path fill="none" stroke="#191970" d="M1022,-895.2455C1022,-887.9382 1022,-877.6944 1022,-868.7046"/>
<polygon fill="#191970" stroke="#191970" points="1025.5001,-868.6426 1022,-858.6427 1018.5001,-868.6427 1025.5001,-868.6426"/>
</g>
diff --git a/docs/reference/api/doxygen/algorithms_8h__incl.svg b/docs/reference/api/doxygen/algorithms_8h__incl.svg
index 9ab219824..959401357 100644
--- a/docs/reference/api/doxygen/algorithms_8h__incl.svg
+++ b/docs/reference/api/doxygen/algorithms_8h__incl.svg
@@ -46,39 +46,39 @@
<path fill="none" stroke="#191970" d="M950.6966,-960.0806C1209.6303,-955.7687 2270.0969,-936.8389 2337,-915 2437.9722,-882.04 2533.0673,-792.0345 2569.9065,-754.1585"/>
<polygon fill="#191970" stroke="#191970" points="2572.8486,-756.1472 2577.2427,-746.5066 2567.7956,-751.3028 2572.8486,-756.1472"/>
</g>
-<!-- Node48 -->
+<!-- Node49 -->
<g id="node41" class="node">
-<title>Node48</title>
+<title>Node49</title>
<g id="a_node41"><a xlink:href="memory__pools_8h.html" target="_top" xlink:title="The object definition for relay.build argument type of memory pools. ">
<polygon fill="#ffffff" stroke="#000000" points="390.5,-895.5 390.5,-914.5 521.5,-914.5 521.5,-895.5 390.5,-895.5"/>
<text text-anchor="middle" x="456" y="-902.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/memory_pools.h</text>
</a>
</g>
</g>
-<!-- Node1->Node48 -->
+<!-- Node1->Node49 -->
<g id="edge124" class="edge">
-<title>Node1->Node48</title>
+<title>Node1->Node49</title>
<path fill="none" stroke="#191970" d="M837.1882,-953.7364C759.1482,-943.7587 618.596,-925.7885 531.6995,-914.6785"/>
<polygon fill="#191970" stroke="#191970" points="532.0005,-911.1885 521.6373,-913.392 531.1127,-918.132 532.0005,-911.1885"/>
</g>
-<!-- Node50 -->
+<!-- Node51 -->
<g id="node43" class="node">
-<title>Node50</title>
+<title>Node51</title>
<g id="a_node43"><a xlink:href="target_8h.html" target="_top" xlink:title="Compilation target object. ">
<polygon fill="#ffffff" stroke="#000000" points="401,-839.5 401,-858.5 511,-858.5 511,-839.5 401,-839.5"/>
<text text-anchor="middle" x="456" y="-846.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/target/target.h</text>
</a>
</g>
</g>
-<!-- Node1->Node50 -->
+<!-- Node1->Node51 -->
<g id="edge160" class="edge">
-<title>Node1->Node50</title>
+<title>Node1->Node51</title>
<path fill="none" stroke="#191970" d="M856.6564,-951.4509C777.0575,-931.0969 590.8262,-883.4761 503.1831,-861.0651"/>
<polygon fill="#191970" stroke="#191970" points="503.8045,-857.6114 493.2491,-858.5249 502.0703,-864.3932 503.8045,-857.6114"/>
</g>
-<!-- Node68 -->
+<!-- Node69 -->
<g id="node48" class="node">
-<title>Node68</title>
+<title>Node69</title>
<g id="a_node48"><a xlink:href="device__api_8h.html" target="_top" xlink:title="Abstract device memory management API. ">
<polygon fill="#ffffff" stroke="#000000" points="1611.5,-492.5 1611.5,-522.5 1724.5,-522.5 1724.5,-492.5 1611.5,-492.5"/>
<text text-anchor="start" x="1619.5" y="-510.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/device</text>
@@ -86,24 +86,24 @@
</a>
</g>
</g>
-<!-- Node1->Node68 -->
+<!-- Node1->Node69 -->
<g id="edge155" class="edge">
-<title>Node1->Node68</title>
+<title>Node1->Node69</title>
<path fill="none" stroke="#191970" d="M950.9003,-959.5732C1216.6771,-952.5422 2323,-919.0671 2323,-849 2323,-849 2323,-849 2323,-793 2323,-727.9193 1882.8457,-577.685 1724.1707,-525.6402"/>
<polygon fill="#191970" stroke="#191970" points="1725.1735,-522.2858 1714.5809,-522.502 1722.9964,-528.9386 1725.1735,-522.2858"/>
</g>
-<!-- Node69 -->
+<!-- Node70 -->
<g id="node49" class="node">
-<title>Node69</title>
+<title>Node70</title>
<g id="a_node49"><a xlink:href="stmt_8h.html" target="_top" xlink:title="TIR statements. ">
<polygon fill="#ffffff" stroke="#000000" points="852,-839.5 852,-858.5 936,-858.5 936,-839.5 852,-839.5"/>
<text text-anchor="middle" x="894" y="-846.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/tir/stmt.h</text>
</a>
</g>
</g>
-<!-- Node1->Node69 -->
+<!-- Node1->Node70 -->
<g id="edge161" class="edge">
-<title>Node1->Node69</title>
+<title>Node1->Node70</title>
<path fill="none" stroke="#191970" d="M894,-951.4509C894,-933.184 894,-892.9553 894,-868.6976"/>
<polygon fill="#191970" stroke="#191970" points="897.5001,-868.5249 894,-858.5249 890.5001,-868.5249 897.5001,-868.5249"/>
</g>
@@ -204,30 +204,30 @@
<path fill="none" stroke="#191970" d="M2613.5697,-727.4257C2651.73,-712.2096 2716,-678.6227 2716,-625 2716,-625 2716,-625 2716,-440.5 2716,-402.8735 2724.0317,-384.1733 2697,-358 2683.4222,-344.8534 2580.8219,-327.5573 2508.1291,-316.7334"/>
<polygon fill="#191970" stroke="#191970" points="2508.4867,-313.2483 2498.0827,-315.2501 2507.4642,-320.1732 2508.4867,-313.2483"/>
</g>
-<!-- Node43 -->
+<!-- Node44 -->
<g id="node38" class="node">
-<title>Node43</title>
+<title>Node44</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="2644,-364 2644,-383 2688,-383 2688,-364 2644,-364"/>
<text text-anchor="middle" x="2666" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">limits</text>
</g>
-<!-- Node2->Node43 -->
+<!-- Node2->Node44 -->
<g id="edge121" class="edge">
-<title>Node2->Node43</title>
+<title>Node2->Node44</title>
<path fill="none" stroke="#191970" d="M2602.4478,-727.232C2628.9102,-710.1274 2678,-672.0556 2678,-625 2678,-625 2678,-625 2678,-507.5 2678,-466.8078 2672.4424,-419.4407 2668.8915,-393.3298"/>
<polygon fill="#191970" stroke="#191970" points="2672.3291,-392.6406 2667.4714,-383.225 2665.3972,-393.6148 2672.3291,-392.6406"/>
</g>
-<!-- Node47 -->
+<!-- Node48 -->
<g id="node40" class="node">
-<title>Node47</title>
+<title>Node48</title>
<g id="a_node40"><a xlink:href="ir_2type_8h.html" target="_top" xlink:title="IR/AST nodes for the unified type system in TVM. ">
<polygon fill="#ffffff" stroke="#000000" points="1179,-671.5 1179,-690.5 1259,-690.5 1259,-671.5 1179,-671.5"/>
<text text-anchor="middle" x="1219" y="-678.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/type.h</text>
</a>
</g>
</g>
-<!-- Node2->Node47 -->
+<!-- Node2->Node48 -->
<g id="edge110" class="edge">
-<title>Node2->Node47</title>
+<title>Node2->Node48</title>
<path fill="none" stroke="#191970" d="M2546.1964,-735.3694C2350.1446,-727.338 1487.8468,-692.0135 1269.4233,-683.0656"/>
<polygon fill="#191970" stroke="#191970" points="1269.3506,-679.5598 1259.2157,-682.6475 1269.064,-686.5539 1269.3506,-679.5598"/>
</g>
@@ -369,18 +369,18 @@
<path fill="none" stroke="#191970" d="M1428.2679,-559.3416C1445.5485,-539.1547 1485.9041,-492.012 1510.2747,-463.5427"/>
<polygon fill="#191970" stroke="#191970" points="1513.1427,-465.5745 1516.9869,-455.7016 1507.825,-461.0223 1513.1427,-465.5745"/>
</g>
-<!-- Node45 -->
+<!-- Node46 -->
<g id="node39" class="node">
-<title>Node45</title>
+<title>Node46</title>
<g id="a_node39"><a xlink:href="repr__printer_8h.html" target="_top" xlink:title="Printer class to print repr string of each AST/IR nodes. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1263.5,-498 1263.5,-517 1394.5,-517 1394.5,-498 1263.5,-498"/>
<text text-anchor="middle" x="1329" y="-505" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/node/repr_printer.h</text>
</a>
</g>
</g>
-<!-- Node4->Node45 -->
+<!-- Node4->Node46 -->
<g id="edge98" class="edge">
-<title>Node4->Node45</title>
+<title>Node4->Node46</title>
<path fill="none" stroke="#191970" d="M1405.7813,-559.3906C1391.3566,-549.6421 1368.8024,-534.3994 1351.864,-522.9521"/>
<polygon fill="#191970" stroke="#191970" points="1353.3641,-519.7415 1343.1189,-517.0419 1349.4444,-525.5412 1353.3641,-519.7415"/>
</g>
@@ -1078,396 +1078,396 @@
<path fill="none" stroke="#191970" d="M1646.7305,-425.4534C1624.4559,-408.3229 1586.4348,-379.8819 1552,-358 1530.6789,-344.4512 1505.5713,-330.6684 1486.8213,-320.7749"/>
<polygon fill="#191970" stroke="#191970" points="1488.3944,-317.6479 1477.9112,-316.1133 1485.1494,-323.8503 1488.3944,-317.6479"/>
</g>
-<!-- Node40->Node43 -->
+<!-- Node40->Node44 -->
<g id="edge89" class="edge">
-<title>Node40->Node43</title>
+<title>Node40->Node44</title>
<path fill="none" stroke="#191970" d="M1724.0866,-436.6082C1910.5603,-424.1145 2490.2291,-385.2766 2633.6544,-375.6672"/>
<polygon fill="#191970" stroke="#191970" points="2634.2174,-379.1374 2643.961,-374.9766 2633.7494,-372.153 2634.2174,-379.1374"/>
</g>
-<!-- Node47->Node3 -->
+<!-- Node48->Node3 -->
<g id="edge111" class="edge">
-<title>Node47->Node3</title>
+<title>Node48->Node3</title>
<path fill="none" stroke="#191970" d="M1241.6916,-671.3733C1263.0711,-662.3032 1295.3057,-648.6279 1319.2362,-638.4755"/>
<polygon fill="#191970" stroke="#191970" points="1320.7656,-641.6287 1328.6045,-634.5011 1318.0317,-635.1846 1320.7656,-641.6287"/>
</g>
-<!-- Node47->Node4 -->
+<!-- Node48->Node4 -->
<g id="edge112" class="edge">
-<title>Node47->Node4</title>
+<title>Node48->Node4</title>
<path fill="none" stroke="#191970" d="M1223.5906,-671.3898C1230.9522,-656.9976 1246.7561,-629.9433 1268,-615 1295.2502,-595.8318 1330.6421,-584.5101 1360.4177,-577.8983"/>
<polygon fill="#191970" stroke="#191970" points="1361.2106,-581.3083 1370.2804,-575.8321 1359.7752,-574.4571 1361.2106,-581.3083"/>
</g>
-<!-- Node47->Node20 -->
+<!-- Node48->Node20 -->
<g id="edge113" class="edge">
-<title>Node47->Node20</title>
+<title>Node48->Node20</title>
<path fill="none" stroke="#191970" d="M1180.6854,-671.4271C1134.7854,-657.3915 1064,-626.4355 1064,-569 1064,-569 1064,-569 1064,-507.5 1064,-432.0138 1129.2371,-362.2409 1166.9265,-328.2299"/>
<polygon fill="#191970" stroke="#191970" points="1169.3389,-330.769 1174.5174,-321.526 1164.7051,-325.5222 1169.3389,-330.769"/>
</g>
-<!-- Node47->Node9 -->
+<!-- Node48->Node9 -->
<g id="edge115" class="edge">
-<title>Node47->Node9</title>
+<title>Node48->Node9</title>
<path fill="none" stroke="#191970" d="M1215.7538,-671.2386C1209.6342,-651.9402 1197,-607.4431 1197,-569 1197,-569 1197,-569 1197,-507.5 1197,-438.9427 1192.9819,-412.1718 1235,-358 1345.9814,-214.9173 1426.1663,-226.1456 1601,-179 1680.8512,-157.4674 1775.2845,-145.2206 1838.1247,-138.8645"/>
<polygon fill="#191970" stroke="#191970" points="1838.7107,-142.3237 1848.318,-137.8571 1838.0222,-135.3576 1838.7107,-142.3237"/>
</g>
-<!-- Node47->Node15 -->
+<!-- Node48->Node15 -->
<g id="edge116" class="edge">
-<title>Node47->Node15</title>
+<title>Node48->Node15</title>
<path fill="none" stroke="#191970" d="M1178.9451,-679.7301C1070.9735,-675.9634 775.7963,-663.3036 683,-635 616.3987,-614.6861 612.0847,-580.9364 546,-559 432.775,-521.4156 380.3449,-587.5244 280,-523 243.5441,-499.5579 228,-483.8424 228,-440.5 228,-440.5 228,-440.5 228,-189 228,-140.5473 278.5515,-104.5212 312.3292,-85.8573"/>
<polygon fill="#191970" stroke="#191970" points="314.1693,-88.843 321.3555,-81.0577 310.8828,-82.6624 314.1693,-88.843"/>
</g>
-<!-- Node47->Node27 -->
+<!-- Node48->Node27 -->
<g id="edge114" class="edge">
-<title>Node47->Node27</title>
+<title>Node48->Node27</title>
<path fill="none" stroke="#191970" d="M1229.6681,-671.2469C1248.0466,-654.6494 1285.2225,-622.0217 1301,-615 1418.4216,-562.7418 1759.9383,-585.9361 1872,-523 1958.7825,-474.2612 2023.4821,-368.2898 2047.1623,-325.2425"/>
<polygon fill="#191970" stroke="#191970" points="2050.273,-326.848 2051.9404,-316.3852 2044.1122,-323.5246 2050.273,-326.848"/>
</g>
-<!-- Node49 -->
+<!-- Node50 -->
<g id="node42" class="node">
-<title>Node49</title>
+<title>Node50</title>
<g id="a_node42"><a xlink:href="registry_8h.html" target="_top" xlink:title="This file defines the TVM global function registry. ">
<polygon fill="#ffffff" stroke="#000000" points="416.5,-498 416.5,-517 541.5,-517 541.5,-498 416.5,-498"/>
<text text-anchor="middle" x="479" y="-505" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/registry.h</text>
</a>
</g>
</g>
-<!-- Node48->Node49 -->
+<!-- Node49->Node50 -->
<g id="edge125" class="edge">
-<title>Node48->Node49</title>
+<title>Node49->Node50</title>
<path fill="none" stroke="#191970" d="M419.205,-895.4434C374.2202,-881.2795 304,-850.0354 304,-793 304,-793 304,-793 304,-625 304,-566.2114 372.9129,-534.9354 424.2495,-519.7925"/>
<polygon fill="#191970" stroke="#191970" points="425.258,-523.1447 433.9332,-517.0627 423.3586,-516.4073 425.258,-523.1447"/>
</g>
-<!-- Node48->Node50 -->
+<!-- Node49->Node51 -->
<g id="edge131" class="edge">
-<title>Node48->Node50</title>
+<title>Node49->Node51</title>
<path fill="none" stroke="#191970" d="M456,-895.2455C456,-887.9382 456,-877.6944 456,-868.7046"/>
<polygon fill="#191970" stroke="#191970" points="459.5001,-868.6426 456,-858.6427 452.5001,-868.6427 459.5001,-868.6426"/>
</g>
-<!-- Node49->Node17 -->
+<!-- Node50->Node17 -->
<g id="edge129" class="edge">
-<title>Node49->Node17</title>
+<title>Node50->Node17</title>
<path fill="none" stroke="#191970" d="M475.6563,-497.8177C465.0902,-465.3454 435.6545,-358.5584 480,-291 501.8412,-257.726 527.0645,-274.0711 562,-255 655.1565,-204.1466 661.2245,-159.1807 761,-123 841.1379,-93.9403 1102.0967,-77.9574 1194.3374,-73.1063"/>
<polygon fill="#191970" stroke="#191970" points="1194.6726,-76.5937 1204.4787,-72.5823 1194.3114,-69.603 1194.6726,-76.5937"/>
</g>
-<!-- Node49->Node19 -->
+<!-- Node50->Node19 -->
<g id="edge130" class="edge">
-<title>Node49->Node19</title>
+<title>Node50->Node19</title>
<path fill="none" stroke="#191970" d="M482.6977,-497.6803C496.3652,-462.8534 547.6478,-344.4765 632,-291 669.2856,-267.3622 797.3296,-253.3147 859.1155,-247.7619"/>
<polygon fill="#191970" stroke="#191970" points="859.4867,-251.2428 869.143,-246.8832 858.8756,-244.2696 859.4867,-251.2428"/>
</g>
-<!-- Node49->Node15 -->
+<!-- Node50->Node15 -->
<g id="edge127" class="edge">
-<title>Node49->Node15</title>
+<title>Node50->Node15</title>
<path fill="none" stroke="#191970" d="M466.4498,-497.8534C443.1786,-479.307 393.2321,-436.3016 366,-389 346.9473,-355.9059 342,-344.6867 342,-306.5 342,-306.5 342,-306.5 342,-189 342,-154.6399 342,-114.628 342,-91.2764"/>
<polygon fill="#191970" stroke="#191970" points="345.5001,-91.2489 342,-81.2489 338.5001,-91.249 345.5001,-91.2489"/>
</g>
-<!-- Node49->Node16 -->
+<!-- Node50->Node16 -->
<g id="edge128" class="edge">
-<title>Node49->Node16</title>
+<title>Node50->Node16</title>
<path fill="none" stroke="#191970" d="M541.7668,-507.0239C725.3701,-505.1795 1277.2155,-496.416 1733,-456 2053.3724,-427.5914 2130.5142,-395.0722 2450,-358 2578.2711,-343.1158 3020,-374.1318 3020,-245 3020,-245 3020,-245 3020,-189 3020,-150.4693 2995.1574,-111.14 2978.5649,-89.1688"/>
<polygon fill="#191970" stroke="#191970" points="2981.1848,-86.8376 2972.263,-81.1234 2975.6741,-91.1541 2981.1848,-86.8376"/>
</g>
-<!-- Node49->Node40 -->
+<!-- Node50->Node40 -->
<g id="edge126" class="edge">
-<title>Node49->Node40</title>
+<title>Node50->Node40</title>
<path fill="none" stroke="#191970" d="M541.5707,-506.0043C745.3812,-500.9134 1388.8717,-483.0589 1595,-456 1595.903,-455.8815 1596.8116,-455.7573 1597.7246,-455.6279"/>
<polygon fill="#191970" stroke="#191970" points="1598.3024,-459.0801 1607.6325,-454.0601 1597.2083,-452.1661 1598.3024,-459.0801"/>
</g>
-<!-- Node50->Node2 -->
+<!-- Node51->Node2 -->
<g id="edge132" class="edge">
-<title>Node50->Node2</title>
+<title>Node51->Node2</title>
<path fill="none" stroke="#191970" d="M511.1132,-847.5135C586.1725,-845.5032 724.7447,-841.8426 843,-839 1022.8814,-834.676 2287.9158,-848.5051 2462,-803 2500.9143,-792.8279 2541.1553,-768.3683 2564.8615,-752.2529"/>
<polygon fill="#191970" stroke="#191970" points="2566.8824,-755.1107 2573.1002,-746.5323 2562.89,-749.3609 2566.8824,-755.1107"/>
</g>
-<!-- Node50->Node4 -->
+<!-- Node51->Node4 -->
<g id="edge143" class="edge">
-<title>Node50->Node4</title>
+<title>Node51->Node4</title>
<path fill="none" stroke="#191970" d="M511.1103,-843.457C579.6195,-836.1573 699.9781,-821.989 802,-803 1011.4125,-764.0227 1065.7073,-757.7121 1268,-691 1328.5217,-671.0411 1357.4117,-682.4075 1400,-635 1411.3643,-622.3497 1416.2749,-603.4207 1418.3947,-589.0739"/>
<polygon fill="#191970" stroke="#191970" points="1421.9154,-589.0891 1419.5666,-578.758 1414.9602,-588.299 1421.9154,-589.0891"/>
</g>
-<!-- Node50->Node19 -->
+<!-- Node51->Node19 -->
<g id="edge154" class="edge">
-<title>Node50->Node19</title>
+<title>Node51->Node19</title>
<path fill="none" stroke="#191970" d="M443.4501,-839.4082C421.864,-821.7155 380,-781.6472 380,-737 380,-737 380,-737 380,-681 380,-577.8506 482.6679,-601.1421 550,-523 637.1824,-421.8207 623.9047,-355.2396 741,-291 778.8965,-270.2096 827.3838,-257.6316 859.277,-251.0046"/>
<polygon fill="#191970" stroke="#191970" points="860.2953,-254.3705 869.4175,-248.9821 858.9261,-247.5057 860.2953,-254.3705"/>
</g>
-<!-- Node50->Node15 -->
+<!-- Node51->Node15 -->
<g id="edge152" class="edge">
-<title>Node50->Node15</title>
+<title>Node51->Node15</title>
<path fill="none" stroke="#191970" d="M400.9957,-847.0134C279.7137,-841.1364 0,-818.7234 0,-737 0,-737 0,-737 0,-189 0,-148.7538 26.6665,-142.2691 62,-123 104.4408,-99.8549 245.5022,-82.0118 309.6925,-74.8815"/>
<polygon fill="#191970" stroke="#191970" points="310.1074,-78.3571 319.6679,-73.7912 309.3468,-71.3986 310.1074,-78.3571"/>
</g>
-<!-- Node51 -->
+<!-- Node52 -->
<g id="node44" class="node">
-<title>Node51</title>
+<title>Node52</title>
<g id="a_node44"><a xlink:href="ir_2module_8h.html" target="_top" xlink:title="IRModule that holds the functions and type definitions. ">
<polygon fill="#ffffff" stroke="#ff0000" points="698.5,-783.5 698.5,-802.5 793.5,-802.5 793.5,-783.5 698.5,-783.5"/>
<text text-anchor="middle" x="746" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/ir/module.h</text>
</a>
</g>
</g>
-<!-- Node50->Node51 -->
+<!-- Node51->Node52 -->
<g id="edge133" class="edge">
-<title>Node50->Node51</title>
+<title>Node51->Node52</title>
<path fill="none" stroke="#191970" d="M505.5233,-839.4369C556.1318,-829.6642 634.6739,-814.4974 688.1202,-804.1768"/>
<polygon fill="#191970" stroke="#191970" points="688.972,-807.577 698.127,-802.2444 687.6448,-800.704 688.972,-807.577"/>
</g>
-<!-- Node59 -->
+<!-- Node60 -->
<g id="node45" class="node">
-<title>Node59</title>
+<title>Node60</title>
<polygon fill="#ffffff" stroke="#bfbfbf" points="522,-727.5 522,-746.5 608,-746.5 608,-727.5 522,-727.5"/>
<text text-anchor="middle" x="565" y="-734.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">unordered_set</text>
</g>
-<!-- Node50->Node59 -->
+<!-- Node51->Node60 -->
<g id="edge153" class="edge">
-<title>Node50->Node59</title>
+<title>Node51->Node60</title>
<path fill="none" stroke="#191970" d="M465.2933,-839.4509C483.7949,-820.4401 525.4463,-777.6423 548.6344,-753.816"/>
<polygon fill="#191970" stroke="#191970" points="551.264,-756.1324 555.7302,-746.5249 546.2475,-751.2502 551.264,-756.1324"/>
</g>
-<!-- Node60 -->
+<!-- Node61 -->
<g id="node46" class="node">
-<title>Node60</title>
+<title>Node61</title>
<g id="a_node46"><a xlink:href="with_8h.html" target="_top" xlink:title="RAII wrapper function to enter and exit a context object similar to python's with syntax...">
<polygon fill="#ffffff" stroke="#ff0000" points="97.5,-671.5 97.5,-690.5 206.5,-690.5 206.5,-671.5 97.5,-671.5"/>
<text text-anchor="middle" x="152" y="-678.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/support/with.h</text>
</a>
</g>
</g>
-<!-- Node50->Node60 -->
+<!-- Node51->Node61 -->
<g id="edge144" class="edge">
-<title>Node50->Node60</title>
+<title>Node51->Node61</title>
<path fill="none" stroke="#191970" d="M436.5551,-839.3826C418.2363,-830.2466 390.0844,-816.0113 366,-803 296.8548,-765.6451 216.7067,-719.0538 176.9827,-695.7355"/>
<polygon fill="#191970" stroke="#191970" points="178.6292,-692.6435 168.2347,-690.5928 175.0816,-698.678 178.6292,-692.6435"/>
</g>
-<!-- Node61 -->
+<!-- Node62 -->
<g id="node47" class="node">
-<title>Node61</title>
+<title>Node62</title>
<g id="a_node47"><a xlink:href="target__kind_8h.html" target="_top" xlink:title="Target kind registry. ">
<polygon fill="#ffffff" stroke="#ff0000" points="692,-615.5 692,-634.5 828,-634.5 828,-615.5 692,-615.5"/>
<text text-anchor="middle" x="760" y="-622.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/target/target_kind.h</text>
</a>
</g>
</g>
-<!-- Node50->Node61 -->
+<!-- Node51->Node62 -->
<g id="edge146" class="edge">
-<title>Node50->Node61</title>
+<title>Node51->Node62</title>
<path fill="none" stroke="#191970" d="M457.9811,-839.36C463.1067,-816.7233 478.8573,-759.6251 513,-727 564.9449,-677.364 644.4304,-650.4548 699.1279,-636.9399"/>
<polygon fill="#191970" stroke="#191970" points="700.1285,-640.2994 709.0404,-634.5698 698.5006,-633.4913 700.1285,-640.2994"/>
</g>
-<!-- Node51->Node2 -->
+<!-- Node52->Node2 -->
<g id="edge134" class="edge">
-<title>Node51->Node2</title>
+<title>Node52->Node2</title>
<path fill="none" stroke="#191970" d="M793.6601,-791.679C983.6194,-786.3948 1705.2141,-766.1386 2299,-747 2382.0324,-744.3237 2478.5656,-740.8948 2535.8465,-738.8268"/>
<polygon fill="#191970" stroke="#191970" points="2536.2063,-742.3162 2546.0733,-738.457 2535.9533,-735.3207 2536.2063,-742.3162"/>
</g>
-<!-- Node51->Node20 -->
+<!-- Node52->Node20 -->
<g id="edge136" class="edge">
-<title>Node51->Node20</title>
+<title>Node52->Node20</title>
<path fill="none" stroke="#191970" d="M751.9703,-783.1404C766.8463,-758.4672 806.5561,-691.9493 837,-635 901.7507,-513.8755 866.4503,-443.4217 974,-358 996.4583,-340.1625 1065.2214,-325.8422 1119.7712,-316.8739"/>
<polygon fill="#191970" stroke="#191970" points="1120.461,-320.308 1129.777,-315.262 1119.3476,-313.3971 1120.461,-320.308"/>
</g>
-<!-- Node51->Node17 -->
+<!-- Node52->Node17 -->
<g id="edge141" class="edge">
-<title>Node51->Node17</title>
+<title>Node52->Node17</title>
<path fill="none" stroke="#191970" d="M732.698,-783.4701C672.2188,-740.0043 425.3341,-561.0409 407,-523 375.8175,-458.3001 329.3536,-475.2302 462,-291 540.436,-182.0617 586.6654,-168.3798 713,-123 802.2852,-90.9285 1095.0568,-76.6927 1193.9937,-72.7172"/>
<polygon fill="#191970" stroke="#191970" points="1194.4274,-76.203 1204.2822,-72.313 1194.1525,-69.2084 1194.4274,-76.203"/>
</g>
-<!-- Node51->Node19 -->
+<!-- Node52->Node19 -->
<g id="edge142" class="edge">
-<title>Node51->Node19</title>
+<title>Node52->Node19</title>
<path fill="none" stroke="#191970" d="M737.98,-783.3001C703.2724,-740.0862 570.4827,-561.2416 632,-425 641.2654,-404.4801 764.9356,-302.9775 784,-291 807.8299,-276.0285 837.3753,-263.906 859.6919,-255.8715"/>
<polygon fill="#191970" stroke="#191970" points="861.111,-259.0829 869.3903,-252.4721 858.7955,-252.477 861.111,-259.0829"/>
</g>
-<!-- Node51->Node15 -->
+<!-- Node52->Node15 -->
<g id="edge138" class="edge">
-<title>Node51->Node15</title>
+<title>Node52->Node15</title>
<path fill="none" stroke="#191970" d="M727.9753,-783.4508C695.2467,-766.1604 628.3684,-731.0709 617,-727 546.3354,-701.6959 519.8379,-723.1215 452,-691 402.8164,-667.7114 402.8287,-642.721 356,-615 310.7519,-588.2147 172.9583,-562.278 138,-523 112.6121,-494.475 114,-478.6867 114,-440.5 114,-440.5 114,-440.5 114,-189 114,-157.7875 114.8384,-143.9227 138,-123 163.1161,-100.3117 258.7714,-83.4839 309.6777,-75.9212"/>
<polygon fill="#191970" stroke="#191970" points="310.4263,-79.3493 319.8199,-74.4492 309.4208,-72.4219 310.4263,-79.3493"/>
</g>
-<!-- Node51->Node32 -->
+<!-- Node52->Node32 -->
<g id="edge137" class="edge">
-<title>Node51->Node32</title>
+<title>Node52->Node32</title>
<path fill="none" stroke="#191970" d="M793.5196,-792.7136C1030.8279,-791.1232 2081.7948,-782.135 2223,-747 2277.7794,-733.3697 2299.7169,-733.3856 2337,-691 2430.6633,-584.5181 2416.7985,-528.9032 2440,-389 2443.1775,-369.8398 2441.5973,-347.8781 2439.4021,-331.4333"/>
<polygon fill="#191970" stroke="#191970" points="2442.8582,-330.8791 2437.9146,-321.5083 2435.9355,-331.9167 2442.8582,-330.8791"/>
</g>
-<!-- Node51->Node35 -->
+<!-- Node52->Node35 -->
<g id="edge139" class="edge">
-<title>Node51->Node35</title>
+<title>Node52->Node35</title>
<path fill="none" stroke="#191970" d="M793.7001,-787.5775C949.851,-769.0456 1458.889,-702.4415 1863,-579 2029.2963,-528.2024 2124.0535,-542.8913 2205,-389 2225.6942,-349.6573 2249.2565,-328.2521 2225,-291 2217.5251,-279.5204 2187.5941,-266.9695 2161.7493,-257.882"/>
<polygon fill="#191970" stroke="#191970" points="2162.694,-254.506 2152.0994,-254.5781 2160.4266,-261.1286 2162.694,-254.506"/>
</g>
-<!-- Node51->Node47 -->
+<!-- Node52->Node48 -->
<g id="edge135" class="edge">
-<title>Node51->Node47</title>
+<title>Node52->Node48</title>
<path fill="none" stroke="#191970" d="M783.3738,-783.4196C834.7832,-770.3117 930.2493,-746.2224 1012,-727 1065.4679,-714.4279 1126.9231,-700.8855 1168.8945,-691.7741"/>
<polygon fill="#191970" stroke="#191970" points="1169.8318,-695.1523 1178.8637,-689.6137 1168.3492,-688.3111 1169.8318,-695.1523"/>
</g>
-<!-- Node51->Node59 -->
+<!-- Node52->Node60 -->
<g id="edge140" class="edge">
-<title>Node51->Node59</title>
+<title>Node52->Node60</title>
<path fill="none" stroke="#191970" d="M714.885,-783.3733C684.6571,-774.021 638.604,-759.7725 605.5261,-749.5385"/>
<polygon fill="#191970" stroke="#191970" points="606.2967,-746.1133 595.709,-746.5011 604.2277,-752.8005 606.2967,-746.1133"/>
</g>
-<!-- Node60->Node17 -->
+<!-- Node61->Node17 -->
<g id="edge145" class="edge">
-<title>Node60->Node17</title>
+<title>Node61->Node17</title>
<path fill="none" stroke="#191970" d="M129.749,-671.3895C96.5729,-655.3825 38,-619.9125 38,-569 38,-569 38,-569 38,-373.5 38,-96.375 340.3727,-191.1001 609,-123 721.4876,-94.4831 1082.7481,-77.4747 1194.0458,-72.8139"/>
<polygon fill="#191970" stroke="#191970" points="1194.4676,-76.2996 1204.3144,-72.3894 1194.1784,-69.3055 1194.4676,-76.2996"/>
</g>
-<!-- Node61->Node4 -->
+<!-- Node62->Node4 -->
<g id="edge147" class="edge">
-<title>Node61->Node4</title>
+<title>Node62->Node4</title>
<path fill="none" stroke="#191970" d="M828.1457,-619.2179C957.7263,-608.2232 1237.6447,-584.4726 1360.3342,-574.0626"/>
<polygon fill="#191970" stroke="#191970" points="1360.6337,-577.5498 1370.3019,-573.2168 1360.0418,-570.5749 1360.6337,-577.5498"/>
</g>
-<!-- Node61->Node22 -->
+<!-- Node62->Node22 -->
<g id="edge148" class="edge">
-<title>Node61->Node22</title>
+<title>Node62->Node22</title>
<path fill="none" stroke="#191970" d="M763.6561,-615.3772C778.7049,-577.0269 839.646,-433.514 936,-358 1013.6646,-297.1331 1128.0496,-265.6871 1187.6399,-252.6006"/>
<polygon fill="#191970" stroke="#191970" points="1188.7125,-255.9506 1197.7612,-250.4396 1187.2508,-249.1049 1188.7125,-255.9506"/>
</g>
-<!-- Node61->Node17 -->
+<!-- Node62->Node17 -->
<g id="edge150" class="edge">
-<title>Node61->Node17</title>
+<title>Node62->Node17</title>
<path fill="none" stroke="#191970" d="M756.0268,-615.3359C743.2269,-583.6125 703.1398,-479.5074 690,-389 687.0025,-368.3534 691.1314,-297.1449 693,-291 701.4711,-263.1424 709.6756,-258.4023 727,-235 769.2609,-177.9127 776.3125,-152.3336 841,-123 903.9783,-94.4414 1112.7812,-78.6397 1194.1427,-73.4429"/>
<polygon fill="#191970" stroke="#191970" points="1194.5443,-76.9247 1204.3061,-72.8068 1194.107,-69.9384 1194.5443,-76.9247"/>
</g>
-<!-- Node61->Node19 -->
+<!-- Node62->Node19 -->
<g id="edge151" class="edge">
-<title>Node61->Node19</title>
+<title>Node62->Node19</title>
<path fill="none" stroke="#191970" d="M759.7349,-615.0998C759.2416,-585.3926 759.9521,-494.761 784,-425 806.3229,-360.2429 854.4149,-293.7894 878.5559,-262.8599"/>
<polygon fill="#191970" stroke="#191970" points="881.4938,-264.7876 884.957,-254.7748 876.0056,-260.4425 881.4938,-264.7876"/>
</g>
-<!-- Node61->Node35 -->
+<!-- Node62->Node35 -->
<g id="edge149" class="edge">
-<title>Node61->Node35</title>
+<title>Node62->Node35</title>
<path fill="none" stroke="#191970" d="M828.3504,-622.18C1047.6125,-612.6542 1731.4286,-579.0942 1945,-523 2061.287,-492.4574 2111.856,-488.055 2180,-389 2205.5403,-351.8744 2229.0667,-329.711 2206,-291 2197.109,-276.0789 2181.5103,-265.7035 2166.0823,-258.6373"/>
<polygon fill="#191970" stroke="#191970" points="2167.1358,-255.2841 2156.5598,-254.6541 2164.4346,-261.742 2167.1358,-255.2841"/>
</g>
-<!-- Node68->Node10 -->
+<!-- Node69->Node10 -->
<g id="edge156" class="edge">
-<title>Node68->Node10</title>
+<title>Node69->Node10</title>
<path fill="none" stroke="#191970" d="M1719.3925,-492.4741C1826.8637,-461.7085 2083.6674,-391.6171 2304,-358 2358.7335,-349.6491 2761.2092,-360.4992 2801,-322 2864.6982,-260.3694 2831.7702,-209.8333 2814,-123 2812.1226,-113.826 2808.5439,-104.2655 2804.8098,-95.9078"/>
<polygon fill="#191970" stroke="#191970" points="2807.906,-94.269 2800.4474,-86.7444 2801.5857,-97.2779 2807.906,-94.269"/>
</g>
-<!-- Node68->Node15 -->
+<!-- Node69->Node15 -->
<g id="edge159" class="edge">
-<title>Node68->Node15</title>
+<title>Node69->Node15</title>
<path fill="none" stroke="#191970" d="M1611.4106,-505.2172C1520.313,-500.712 1335.8797,-488.1561 1183,-456 835.2031,-382.8458 460.0209,-148.6811 364.9018,-86.6816"/>
<polygon fill="#191970" stroke="#191970" points="366.6251,-83.6263 356.3425,-81.0729 362.7885,-89.4813 366.6251,-83.6263"/>
</g>
-<!-- Node68->Node29 -->
+<!-- Node69->Node29 -->
<g id="edge157" class="edge">
-<title>Node68->Node29</title>
+<title>Node69->Node29</title>
<path fill="none" stroke="#191970" d="M1689.0634,-492.4765C1702.2279,-482.6914 1719.2345,-469.3108 1733,-456 1754.3881,-435.3184 1776.1064,-408.6472 1789.5588,-391.3152"/>
<polygon fill="#191970" stroke="#191970" points="1792.4864,-393.2493 1795.7946,-383.1842 1786.9317,-388.9894 1792.4864,-393.2493"/>
</g>
-<!-- Node68->Node40 -->
+<!-- Node69->Node40 -->
<g id="edge158" class="edge">
-<title>Node68->Node40</title>
+<title>Node69->Node40</title>
<path fill="none" stroke="#191970" d="M1667.5462,-492.2967C1667.3135,-484.5013 1667.0239,-474.7991 1666.7584,-465.9064"/>
<polygon fill="#191970" stroke="#191970" points="1670.2489,-465.5343 1666.452,-455.6432 1663.2521,-465.7432 1670.2489,-465.5343"/>
</g>
-<!-- Node69->Node17 -->
+<!-- Node70->Node17 -->
<g id="edge176" class="edge">
-<title>Node69->Node17</title>
+<title>Node70->Node17</title>
<path fill="none" stroke="#191970" d="M894,-839.4116C894,-820.0538 894,-774.9167 894,-737 894,-737 894,-737 894,-569 894,-419.7884 789.2351,-366.3637 860,-235 926.9804,-110.6615 1117.5888,-80.7284 1194.3642,-73.647"/>
<polygon fill="#191970" stroke="#191970" points="1194.7499,-77.1269 1204.4173,-72.792 1194.1567,-70.1521 1194.7499,-77.1269"/>
</g>
-<!-- Node69->Node19 -->
+<!-- Node70->Node19 -->
<g id="edge177" class="edge">
-<title>Node69->Node19</title>
+<title>Node70->Node19</title>
<path fill="none" stroke="#191970" d="M851.7272,-844.2614C761.1328,-833.1262 553.2156,-801.9981 513,-747 476.4648,-697.0351 789.6924,-335.224 833,-291 844.1237,-279.641 858.1531,-268.733 869.8483,-260.3922"/>
<polygon fill="#191970" stroke="#191970" points="872.0384,-263.1324 878.25,-254.5494 868.0418,-257.3854 872.0384,-263.1324"/>
</g>
-<!-- Node69->Node15 -->
+<!-- Node70->Node15 -->
<g id="edge174" class="edge">
-<title>Node69->Node15</title>
+<title>Node70->Node15</title>
<path fill="none" stroke="#191970" d="M851.7671,-845.2646C775.2902,-838.1429 616.9746,-821.6591 565,-803 437.687,-757.294 76,-575.7688 76,-440.5 76,-440.5 76,-440.5 76,-189 76,-154.2309 89.602,-143.061 118,-123 148.7073,-101.3076 255.5759,-83.6463 309.8457,-75.8351"/>
<polygon fill="#191970" stroke="#191970" points="310.5483,-79.2707 319.9603,-74.4059 309.5689,-72.3395 310.5483,-79.2707"/>
</g>
-<!-- Node69->Node16 -->
+<!-- Node70->Node16 -->
<g id="edge175" class="edge">
-<title>Node69->Node16</title>
+<title>Node70->Node16</title>
<path fill="none" stroke="#191970" d="M936.2469,-848.6483C1121.3903,-846.9261 1879.575,-837.9909 2501,-803 2675.7831,-793.1584 3286,-912.0599 3286,-737 3286,-737 3286,-737 3286,-189 3286,-130.6847 3097.2688,-92.7908 3008.7863,-78.2296"/>
<polygon fill="#191970" stroke="#191970" points="3009.196,-74.7504 2998.7656,-76.6098 3008.079,-81.6607 3009.196,-74.7504"/>
</g>
-<!-- Node70 -->
+<!-- Node71 -->
<g id="node50" class="node">
-<title>Node70</title>
+<title>Node71</title>
<g id="a_node50"><a xlink:href="tir_2expr_8h.html" target="_top" xlink:title="TIR expressions. ">
<polygon fill="#ffffff" stroke="#ff0000" points="2369.5,-783.5 2369.5,-802.5 2452.5,-802.5 2452.5,-783.5 2369.5,-783.5"/>
<text text-anchor="middle" x="2411" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/tir/expr.h</text>
</a>
</g>
</g>
-<!-- Node69->Node70 -->
+<!-- Node70->Node71 -->
<g id="edge162" class="edge">
-<title>Node69->Node70</title>
+<title>Node70->Node71</title>
<path fill="none" stroke="#191970" d="M936.1566,-847.4438C1151.0512,-839.511 2123.5761,-803.6102 2359.3813,-794.9055"/>
<polygon fill="#191970" stroke="#191970" points="2359.5405,-798.4021 2369.4046,-794.5355 2359.2822,-791.4069 2359.5405,-798.4021"/>
</g>
-<!-- Node70->Node2 -->
+<!-- Node71->Node2 -->
<g id="edge163" class="edge">
-<title>Node70->Node2</title>
+<title>Node71->Node2</title>
<path fill="none" stroke="#191970" d="M2441.0836,-783.3733C2470.1834,-774.0613 2514.4523,-759.8953 2546.403,-749.671"/>
<polygon fill="#191970" stroke="#191970" points="2547.8515,-752.8824 2556.309,-746.5011 2545.718,-746.2155 2547.8515,-752.8824"/>
</g>
-<!-- Node70->Node4 -->
+<!-- Node71->Node4 -->
<g id="edge164" class="edge">
-<title>Node70->Node4</title>
+<title>Node71->Node4</title>
<path fill="none" stroke="#191970" d="M2369.3658,-783.5892C2211.1541,-747.828 1647.3887,-620.3977 1472.2771,-580.8164"/>
<polygon fill="#191970" stroke="#191970" points="1472.8069,-577.3479 1462.2812,-578.557 1471.2635,-584.1757 1472.8069,-577.3479"/>
</g>
-<!-- Node70->Node20 -->
+<!-- Node71->Node20 -->
<g id="edge166" class="edge">
-<title>Node70->Node20</title>
+<title>Node71->Node20</title>
<path fill="none" stroke="#191970" d="M2369.3881,-792.4859C2147.5188,-789.6327 1111.5141,-774.856 1050,-747 1013.3377,-730.3979 988,-721.2462 988,-681 988,-681 988,-681 988,-440.5 988,-375.4528 1061.2087,-340.2954 1120.1283,-322.5334"/>
<polygon fill="#191970" stroke="#191970" points="1121.1888,-325.8702 1129.8205,-319.7266 1119.2416,-319.1465 1121.1888,-325.8702"/>
</g>
-<!-- Node70->Node21 -->
+<!-- Node71->Node21 -->
<g id="edge169" class="edge">
-<title>Node70->Node21</title>
+<title>Node71->Node21</title>
<path fill="none" stroke="#191970" d="M2452.5614,-791.3921C2521.8743,-787.9275 2665.0097,-777.4248 2782,-747 2857.2432,-727.4321 2944,-758.746 2944,-681 2944,-681 2944,-681 2944,-625 2944,-454.873 2887.5447,-400.0931 2757,-291 2677.4919,-224.557 2553.4588,-201.0168 2488.6709,-192.9714"/>
<polygon fill="#191970" stroke="#191970" points="2488.7595,-189.4578 2478.4194,-191.767 2487.9427,-196.41 2488.7595,-189.4578"/>
</g>
-<!-- Node70->Node17 -->
+<!-- Node71->Node17 -->
<g id="edge173" class="edge">
-<title>Node70->Node17</title>
+<title>Node71->Node17</title>
<path fill="none" stroke="#191970" d="M2369.405,-792.2002C2180.2165,-788.1648 1400.266,-767.1297 1170,-691 966.7301,-623.7956 988,-459.0912 988,-245 988,-245 988,-245 988,-189 988,-143.5754 1128.7963,-98.6606 1194.2575,-80.2329"/>
<polygon fill="#191970" stroke="#191970" points="1195.3834,-83.5528 1204.0837,-77.5067 1193.5119,-76.8076 1195.3834,-83.5528"/>
</g>
-<!-- Node70->Node10 -->
+<!-- Node71->Node10 -->
<g id="edge165" class="edge">
-<title>Node70->Node10</title>
+<title>Node71->Node10</title>
<path fill="none" stroke="#191970" d="M2452.5536,-791.6009C2616.5607,-785.5371 3210,-758.2582 3210,-681 3210,-681 3210,-681 3210,-373.5 3210,-231.0348 3136.1488,-189.2029 3010,-123 2965.4345,-99.612 2910.2135,-86.7093 2866.5004,-79.6676"/>
<polygon fill="#191970" stroke="#191970" points="2866.9219,-76.1913 2856.5053,-78.1266 2865.8552,-83.1096 2866.9219,-76.1913"/>
</g>
-<!-- Node70->Node15 -->
+<!-- Node71->Node15 -->
<g id="edge171" class="edge">
-<title>Node70->Node15</title>
+<title>Node71->Node15</title>
<path fill="none" stroke="#191970" d="M2369.267,-792.7592C2166.2188,-791.4158 1280.3075,-783.5283 1007,-747 900.1488,-732.719 875.1903,-718.6681 771,-691 643.1636,-657.0526 304.0832,-604.6165 200,-523 166.6181,-496.8236 152,-482.9212 152,-440.5 152,-440.5 152,-440.5 152,-189 152,-115.7342 255.4585,-86.6153 309.6884,-76.3846"/>
<polygon fill="#191970" stroke="#191970" points="310.5989,-79.7778 319.8327,-74.5828 309.3747,-72.8856 310.5989,-79.7778"/>
</g>
-<!-- Node70->Node27 -->
+<!-- Node71->Node27 -->
<g id="edge168" class="edge">
-<title>Node70->Node27</title>
+<title>Node71->Node27</title>
<path fill="none" stroke="#191970" d="M2409.2807,-783.4486C2405.9626,-764.1572 2399,-719.1337 2399,-681 2399,-681 2399,-681 2399,-625 2399,-458.9855 2180.5168,-354.8829 2092.8065,-319.7796"/>
<polygon fill="#191970" stroke="#191970" points="2093.8555,-316.4313 2083.2683,-316.0289 2091.2938,-322.9457 2093.8555,-316.4313"/>
</g>
-<!-- Node70->Node32 -->
+<!-- Node71->Node32 -->
<g id="edge167" class="edge">
-<title>Node70->Node32</title>
+<title>Node71->Node32</title>
<path fill="none" stroke="#191970" d="M2452.7072,-790.7271C2558.6883,-783.8177 2830,-758.5559 2830,-681 2830,-681 2830,-681 2830,-440.5 2830,-379.3889 2776.0185,-382.4232 2720,-358 2682.9341,-341.8399 2579.9438,-325.657 2508.4985,-315.859"/>
<polygon fill="#191970" stroke="#191970" points="2508.5779,-312.3377 2498.1979,-314.4604 2507.6361,-319.2741 2508.5779,-312.3377"/>
</g>
-<!-- Node70->Node35 -->
+<!-- Node71->Node35 -->
<g id="edge172" class="edge">
-<title>Node70->Node35</title>
+<title>Node71->Node35</title>
<path fill="none" stroke="#191970" d="M2414.8364,-783.2883C2422.0687,-764.0769 2437,-719.7253 2437,-681 2437,-681 2437,-681 2437,-625 2437,-467.7112 2449.0618,-380.9044 2320,-291 2313.9957,-286.8174 2232.9787,-268.8996 2176.1935,-256.701"/>
<polygon fill="#191970" stroke="#191970" points="2176.7601,-253.243 2166.2486,-254.5697 2175.2932,-260.0876 2176.7601,-253.243"/>
</g>
-<!-- Node70->Node43 -->
+<!-- Node71->Node44 -->
<g id="edge170" class="edge">
-<title>Node70->Node43</title>
+<title>Node71->Node44</title>
<path fill="none" stroke="#191970" d="M2420.6043,-783.0839C2452.7317,-749.4149 2557.7002,-635.3084 2616,-523 2638.9206,-478.8458 2654.5007,-422.0192 2661.6083,-392.7791"/>
<polygon fill="#191970" stroke="#191970" points="2665.0239,-393.5439 2663.9151,-383.0072 2658.2111,-391.9355 2665.0239,-393.5439"/>
</g>
diff --git a/docs/reference/api/doxygen/analyzer_8h__dep__incl.svg b/docs/reference/api/doxygen/analyzer_8h__dep__incl.svg
index 3ea2ecba0..36f7f4d14 100644
--- a/docs/reference/api/doxygen/analyzer_8h__dep__incl.svg
+++ b/docs/reference/api/doxygen/analyzer_8h__dep__incl.svg
@@ -9,15 +9,15 @@
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 560)">
<title>include/tvm/arith/analyzer.h</title>
<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-560 3557.5,-560 3557.5,4 -4,4"/>
-<!-- Node53 -->
+<!-- Node54 -->
<g id="node1" class="node">
-<title>Node53</title>
+<title>Node54</title>
<polygon fill="#bfbfbf" stroke="#000000" points="3171,-536.5 3171,-555.5 3324,-555.5 3324,-536.5 3171,-536.5"/>
<text text-anchor="middle" x="3247.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/arith/analyzer.h</text>
</g>
-<!-- Node54 -->
+<!-- Node55 -->
<g id="node2" class="node">
-<title>Node54</title>
+<title>Node55</title>
<g id="a_node2"><a xlink:href="int__solver_8h.html" target="_top" xlink:title="integer constraints data structures and solvers ">
<polygon fill="#ffffff" stroke="#000000" points="1604.5,-469.5 1604.5,-499.5 1720.5,-499.5 1720.5,-469.5 1604.5,-469.5"/>
<text text-anchor="start" x="1612.5" y="-487.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/arith/int</text>
@@ -25,15 +25,15 @@
</a>
</g>
</g>
-<!-- Node53->Node54 -->
+<!-- Node54->Node55 -->
<g id="edge1" class="edge">
-<title>Node53->Node54</title>
+<title>Node54->Node55</title>
<path fill="none" stroke="#191970" d="M3160.8458,-545.6179C2881.911,-544.0348 2009.2562,-536.2385 1729.5,-500 1726.6583,-499.6319 1723.7606,-499.1966 1720.8423,-498.7099"/>
<polygon fill="#191970" stroke="#191970" points="3160.8933,-549.1181 3170.9126,-545.6738 3160.9322,-542.1182 3160.8933,-549.1181"/>
</g>
-<!-- Node55 -->
+<!-- Node56 -->
<g id="node3" class="node">
-<title>Node55</title>
+<title>Node56</title>
<g id="a_node3"><a xlink:href="iter__affine__map_8h.html" target="_top" xlink:title="Iterator quasi-affine mapping patterns. ">
<polygon fill="#ffffff" stroke="#000000" points="1739,-469.5 1739,-499.5 1858,-499.5 1858,-469.5 1739,-469.5"/>
<text text-anchor="start" x="1747" y="-487.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/arith/iter</text>
@@ -41,30 +41,30 @@
</a>
</g>
</g>
-<!-- Node53->Node55 -->
+<!-- Node54->Node56 -->
<g id="edge2" class="edge">
-<title>Node53->Node55</title>
+<title>Node54->Node56</title>
<path fill="none" stroke="#191970" d="M3160.3362,-545.2337C2898.1187,-542.59 2119.1585,-532.1707 1867.5,-500 1864.4554,-499.6108 1861.3473,-499.1498 1858.2173,-498.6347"/>
<polygon fill="#191970" stroke="#191970" points="3160.5321,-548.7357 3170.5665,-545.3355 3160.6018,-541.7361 3160.5321,-548.7357"/>
</g>
-<!-- Node56 -->
+<!-- Node57 -->
<g id="node4" class="node">
-<title>Node56</title>
+<title>Node57</title>
<g id="a_node4"><a xlink:href="operation_8h.html" target="_top" xlink:title="Operation node can generate one or multiple Tensors. ">
<polygon fill="#ffffff" stroke="#000000" points="1876.5,-475 1876.5,-494 2022.5,-494 2022.5,-475 1876.5,-475"/>
<text text-anchor="middle" x="1949.5" y="-482" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/te/operation.h</text>
</a>
</g>
</g>
-<!-- Node53->Node56 -->
+<!-- Node54->Node57 -->
<g id="edge3" class="edge">
-<title>Node53->Node56</title>
+<title>Node54->Node57</title>
<path fill="none" stroke="#191970" d="M3160.5605,-541.8808C2919.6513,-530.4663 2246.5657,-498.5751 2022.8461,-487.9752"/>
<polygon fill="#191970" stroke="#191970" points="3160.5169,-545.3825 3170.6714,-542.3598 3160.8483,-538.3904 3160.5169,-545.3825"/>
</g>
-<!-- Node74 -->
+<!-- Node75 -->
<g id="node22" class="node">
-<title>Node74</title>
+<title>Node75</title>
<g id="a_node22"><a xlink:href="nn_2pooling_8h.html" target="_top" xlink:title="Pooling op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="2953,-.5 2953,-30.5 3064,-30.5 3064,-.5 2953,-.5"/>
<text text-anchor="start" x="2961" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -72,30 +72,30 @@
</a>
</g>
</g>
-<!-- Node53->Node74 -->
+<!-- Node54->Node75 -->
<g id="edge112" class="edge">
-<title>Node53->Node74</title>
+<title>Node54->Node75</title>
<path fill="none" stroke="#191970" d="M3288.9226,-532.9752C3336.33,-515.1774 3407.5,-478.3289 3407.5,-417.5 3407.5,-417.5 3407.5,-417.5 3407.5,-149.5 3407.5,-77.9798 3174.2098,-37.4872 3064.3994,-22.3886"/>
<polygon fill="#191970" stroke="#191970" points="3287.7081,-529.6925 3279.4938,-536.384 3290.0881,-536.2755 3287.7081,-529.6925"/>
</g>
-<!-- Node76 -->
+<!-- Node77 -->
<g id="node24" class="node">
-<title>Node76</title>
+<title>Node77</title>
<g id="a_node24"><a xlink:href="topi_2nn_8h.html" target="_top" xlink:title="NN op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="3077.5,-73 3077.5,-92 3197.5,-92 3197.5,-73 3077.5,-73"/>
<text text-anchor="middle" x="3137.5" y="-80" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn.h</text>
</a>
</g>
</g>
-<!-- Node53->Node76 -->
+<!-- Node54->Node77 -->
<g id="edge113" class="edge">
-<title>Node53->Node76</title>
+<title>Node54->Node77</title>
<path fill="none" stroke="#191970" d="M3276.7754,-531.5131C3312.8705,-511.3449 3369.5,-471.1535 3369.5,-417.5 3369.5,-417.5 3369.5,-417.5 3369.5,-216.5 3369.5,-137.0782 3266.2882,-103.9901 3197.5475,-90.7603"/>
<polygon fill="#191970" stroke="#191970" points="3274.7869,-528.6088 3267.6458,-536.4355 3278.109,-534.7703 3274.7869,-528.6088"/>
</g>
-<!-- Node80 -->
+<!-- Node81 -->
<g id="node28" class="node">
-<title>Node80</title>
+<title>Node81</title>
<g id="a_node28"><a xlink:href="constant__utils_8h.html" target="_top" xlink:title="Utility functions for handling constants in TVM expressions. ">
<polygon fill="#ffffff" stroke="#000000" points="2921,-402.5 2921,-432.5 3048,-432.5 3048,-402.5 2921,-402.5"/>
<text text-anchor="start" x="2929" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -103,15 +103,15 @@
</a>
</g>
</g>
-<!-- Node53->Node80 -->
+<!-- Node54->Node81 -->
<g id="edge109" class="edge">
-<title>Node53->Node80</title>
+<title>Node54->Node81</title>
<path fill="none" stroke="#191970" d="M3218.4705,-531.8164C3168.5617,-507.4313 3066.9223,-457.771 3015.613,-432.7016"/>
<polygon fill="#191970" stroke="#191970" points="3217.2108,-535.0963 3227.7323,-536.3416 3220.2839,-528.8069 3217.2108,-535.0963"/>
</g>
-<!-- Node83 -->
+<!-- Node84 -->
<g id="node31" class="node">
-<title>Node83</title>
+<title>Node84</title>
<g id="a_node31"><a xlink:href="nn_2bnn_8h.html" target="_top" xlink:title="Binary op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="3230,-335.5 3230,-365.5 3341,-365.5 3341,-335.5 3230,-335.5"/>
<text text-anchor="start" x="3238" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -119,15 +119,15 @@
</a>
</g>
</g>
-<!-- Node53->Node83 -->
+<!-- Node54->Node84 -->
<g id="edge110" class="edge">
-<title>Node53->Node83</title>
+<title>Node54->Node84</title>
<path fill="none" stroke="#191970" d="M3261.3216,-528.4371C3277.0457,-507.2338 3301.7978,-469.8177 3311.5,-433 3315.0109,-419.6771 3314.6915,-415.403 3311.5,-402 3308.3911,-388.9439 3301.4132,-375.4455 3295.4684,-365.5452"/>
<polygon fill="#191970" stroke="#191970" points="3258.471,-526.4035 3255.2056,-536.4826 3264.0437,-530.6397 3258.471,-526.4035"/>
</g>
-<!-- Node97 -->
+<!-- Node98 -->
<g id="node45" class="node">
-<title>Node97</title>
+<title>Node98</title>
<g id="a_node45"><a xlink:href="dilate_8h.html" target="_top" xlink:title="Dilate op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="3192,-402.5 3192,-432.5 3303,-432.5 3303,-402.5 3192,-402.5"/>
<text text-anchor="start" x="3200" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -135,15 +135,15 @@
</a>
</g>
</g>
-<!-- Node53->Node97 -->
+<!-- Node54->Node98 -->
<g id="edge111" class="edge">
-<title>Node53->Node97</title>
+<title>Node54->Node98</title>
<path fill="none" stroke="#191970" d="M3247.5,-526.2718C3247.5,-500.5195 3247.5,-455.9952 3247.5,-432.7016"/>
<polygon fill="#191970" stroke="#191970" points="3244.0001,-526.3416 3247.5,-536.3416 3251.0001,-526.3416 3244.0001,-526.3416"/>
</g>
-<!-- Node100 -->
+<!-- Node101 -->
<g id="node48" class="node">
-<title>Node100</title>
+<title>Node101</title>
<g id="a_node48"><a xlink:href="greedy_8h.html" target="_top" xlink:title="This header file contains helper methods used in greedy algorithms for planning memory for USMP...">
<polygon fill="#ffffff" stroke="#000000" points="3435.5,-469.5 3435.5,-499.5 3553.5,-499.5 3553.5,-469.5 3435.5,-469.5"/>
<text text-anchor="start" x="3443.5" y="-487.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/usmp</text>
@@ -151,15 +151,15 @@
</a>
</g>
</g>
-<!-- Node53->Node100 -->
+<!-- Node54->Node101 -->
<g id="edge108" class="edge">
-<title>Node53->Node100</title>
+<title>Node54->Node101</title>
<path fill="none" stroke="#191970" d="M3295.5975,-534.0243C3335.5664,-524.0725 3392.6053,-509.8705 3435.3613,-499.2248"/>
<polygon fill="#191970" stroke="#191970" points="3294.6808,-530.6456 3285.8227,-536.4581 3296.3721,-537.4382 3294.6808,-530.6456"/>
</g>
-<!-- Node57 -->
+<!-- Node58 -->
<g id="node5" class="node">
-<title>Node57</title>
+<title>Node58</title>
<g id="a_node5"><a xlink:href="cublas_8h.html" target="_top" xlink:title="External function interface to cuBLAS libraries. ">
<polygon fill="#ffffff" stroke="#000000" points="751,-335.5 751,-365.5 884,-365.5 884,-335.5 751,-335.5"/>
<text text-anchor="start" x="759" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/contrib</text>
@@ -167,15 +167,15 @@
</a>
</g>
</g>
-<!-- Node56->Node57 -->
+<!-- Node57->Node58 -->
<g id="edge4" class="edge">
-<title>Node56->Node57</title>
+<title>Node57->Node58</title>
<path fill="none" stroke="#191970" d="M1894.8197,-473.1647C1885.4083,-471.5364 1875.6974,-470.0574 1866.5,-469 1507.6205,-427.7412 1409.2668,-501.0978 1054.5,-433 979.4099,-418.5864 895.8914,-385.1623 850.7911,-365.5669"/>
<polygon fill="#191970" stroke="#191970" points="1894.4136,-476.6478 1904.8759,-474.9773 1895.6553,-469.7588 1894.4136,-476.6478"/>
</g>
-<!-- Node58 -->
+<!-- Node59 -->
<g id="node6" class="node">
-<title>Node58</title>
+<title>Node59</title>
<g id="a_node6"><a xlink:href="cuda_2dense_8h.html" target="_top" xlink:title="CUDA schedule for dense operation. ">
<polygon fill="#ffffff" stroke="#000000" points="659.5,-201.5 659.5,-231.5 781.5,-231.5 781.5,-201.5 659.5,-201.5"/>
<text text-anchor="start" x="667.5" y="-219.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/cuda</text>
@@ -183,15 +183,15 @@
</a>
</g>
</g>
-<!-- Node56->Node58 -->
+<!-- Node57->Node59 -->
<g id="edge9" class="edge">
-<title>Node56->Node58</title>
+<title>Node57->Node59</title>
<path fill="none" stroke="#191970" d="M1895.4621,-473.1778C1885.8531,-471.5168 1875.9098,-470.0235 1866.5,-469 1804.5003,-462.2561 790.2919,-475.3385 744.5,-433 687.1607,-379.9849 706.7098,-271.6282 716.4135,-231.7068"/>
<polygon fill="#191970" stroke="#191970" points="1894.8558,-476.6248 1905.3185,-474.9567 1896.0992,-469.7361 1894.8558,-476.6248"/>
</g>
-<!-- Node59 -->
+<!-- Node60 -->
<g id="node7" class="node">
-<title>Node59</title>
+<title>Node60</title>
<g id="a_node7"><a xlink:href="rocm_2dense_8h.html" target="_top" xlink:title="rocm schedule for dense operation ">
<polygon fill="#ffffff" stroke="#000000" points="612.5,-134.5 612.5,-164.5 736.5,-164.5 736.5,-134.5 612.5,-134.5"/>
<text text-anchor="start" x="620.5" y="-152.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/rocm</text>
@@ -199,15 +199,15 @@
</a>
</g>
</g>
-<!-- Node56->Node59 -->
+<!-- Node57->Node60 -->
<g id="edge98" class="edge">
-<title>Node56->Node59</title>
+<title>Node57->Node60</title>
<path fill="none" stroke="#191970" d="M1997.3621,-471.8076C2021.7406,-463.6225 2050.7436,-451.0707 2072.5,-433 2129.7769,-385.4263 2156.6237,-318.0528 2101.5,-268 1999.1825,-175.0947 997.869,-154.1792 736.5938,-150.2876"/>
<polygon fill="#191970" stroke="#191970" points="1996.0334,-468.558 1987.5801,-474.945 1998.1713,-475.2236 1996.0334,-468.558"/>
</g>
-<!-- Node60 -->
+<!-- Node61 -->
<g id="node8" class="node">
-<title>Node60</title>
+<title>Node61</title>
<g id="a_node8"><a xlink:href="rocblas_8h.html" target="_top" xlink:title="include/tvm/topi/contrib\l/rocblas.h">
<polygon fill="#ffffff" stroke="#000000" points="0,-335.5 0,-365.5 133,-365.5 133,-335.5 0,-335.5"/>
<text text-anchor="start" x="8" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/contrib</text>
@@ -215,15 +215,15 @@
</a>
</g>
</g>
-<!-- Node56->Node60 -->
+<!-- Node57->Node61 -->
<g id="edge7" class="edge">
-<title>Node56->Node60</title>
+<title>Node57->Node61</title>
<path fill="none" stroke="#191970" d="M1895.8027,-473.1469C1886.0901,-471.4723 1876.0235,-469.9823 1866.5,-469 1771.4831,-459.1996 235.1895,-470.8788 147.5,-433 115.069,-418.9909 88.9455,-385.2502 75.7854,-365.5303"/>
<polygon fill="#191970" stroke="#191970" points="1895.2956,-476.6119 1905.7586,-474.9454 1896.5401,-469.7234 1895.2956,-476.6119"/>
</g>
-<!-- Node61 -->
+<!-- Node62 -->
<g id="node9" class="node">
-<title>Node61</title>
+<title>Node62</title>
<g id="a_node9"><a xlink:href="cuda_2injective_8h.html" target="_top" xlink:title="CUDA schedule for injective operations. ">
<polygon fill="#ffffff" stroke="#000000" points="1968.5,-335.5 1968.5,-365.5 2090.5,-365.5 2090.5,-335.5 1968.5,-335.5"/>
<text text-anchor="start" x="1976.5" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/cuda</text>
@@ -231,15 +231,15 @@
</a>
</g>
</g>
-<!-- Node56->Node61 -->
+<!-- Node57->Node62 -->
<g id="edge10" class="edge">
-<title>Node56->Node61</title>
+<title>Node57->Node62</title>
<path fill="none" stroke="#191970" d="M1960.4742,-466.1182C1976.423,-439.404 2005.7096,-390.349 2020.4785,-365.611"/>
<polygon fill="#191970" stroke="#191970" points="1957.3744,-464.4827 1955.2534,-474.8631 1963.3847,-468.071 1957.3744,-464.4827"/>
</g>
-<!-- Node62 -->
+<!-- Node63 -->
<g id="node10" class="node">
-<title>Node62</title>
+<title>Node63</title>
<g id="a_node10"><a xlink:href="rocm_2injective_8h.html" target="_top" xlink:title="rocm schedule for injective operations ">
<polygon fill="#ffffff" stroke="#000000" points="1968.5,-268.5 1968.5,-298.5 2092.5,-298.5 2092.5,-268.5 1968.5,-268.5"/>
<text text-anchor="start" x="1976.5" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/rocm</text>
@@ -247,15 +247,15 @@
</a>
</g>
</g>
-<!-- Node56->Node62 -->
+<!-- Node57->Node63 -->
<g id="edge99" class="edge">
-<title>Node56->Node62</title>
+<title>Node57->Node63</title>
<path fill="none" stroke="#191970" d="M1980.6961,-470.4997C2016.948,-452.2971 2074.9298,-416.7229 2099.5,-366 2105.5064,-353.6004 2105.6759,-347.3161 2099.5,-335 2091.5935,-319.2327 2076.3158,-307.0591 2062.1421,-298.5161"/>
<polygon fill="#191970" stroke="#191970" points="1978.9943,-467.4354 1971.5374,-474.9618 1982.0602,-473.7283 1978.9943,-467.4354"/>
</g>
-<!-- Node63 -->
+<!-- Node64 -->
<g id="node11" class="node">
-<title>Node63</title>
+<title>Node64</title>
<g id="a_node11"><a xlink:href="cuda_2pooling_8h.html" target="_top" xlink:title="CUDA schedule for pooling operations. ">
<polygon fill="#ffffff" stroke="#000000" points="1714.5,-335.5 1714.5,-365.5 1836.5,-365.5 1836.5,-335.5 1714.5,-335.5"/>
<text text-anchor="start" x="1722.5" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/cuda</text>
@@ -263,15 +263,15 @@
</a>
</g>
</g>
-<!-- Node56->Node63 -->
+<!-- Node57->Node64 -->
<g id="edge12" class="edge">
-<title>Node56->Node63</title>
+<title>Node57->Node64</title>
<path fill="none" stroke="#191970" d="M1903.3079,-471.7685C1879.4161,-463.5268 1850.8662,-450.935 1829.5,-433 1807.0795,-414.18 1790.5176,-383.6923 1781.9911,-365.5214"/>
<polygon fill="#191970" stroke="#191970" points="1902.2938,-475.1194 1912.8869,-474.9299 1904.4877,-468.4721 1902.2938,-475.1194"/>
</g>
-<!-- Node64 -->
+<!-- Node65 -->
<g id="node12" class="node">
-<title>Node64</title>
+<title>Node65</title>
<g id="a_node12"><a xlink:href="rocm_2pooling_8h.html" target="_top" xlink:title="rocm schedule for pooling operations ">
<polygon fill="#ffffff" stroke="#000000" points="1676.5,-268.5 1676.5,-298.5 1800.5,-298.5 1800.5,-268.5 1676.5,-268.5"/>
<text text-anchor="start" x="1684.5" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/rocm</text>
@@ -279,15 +279,15 @@
</a>
</g>
</g>
-<!-- Node56->Node64 -->
+<!-- Node57->Node65 -->
<g id="edge100" class="edge">
-<title>Node56->Node64</title>
+<title>Node57->Node65</title>
<path fill="none" stroke="#191970" d="M1902.339,-472.1181C1872.0579,-463.2553 1832.3885,-449.9129 1799.5,-433 1753.8753,-409.5375 1728.9987,-411.606 1705.5,-366 1693.8416,-343.3736 1710.9205,-315.6419 1724.5356,-298.7853"/>
<polygon fill="#191970" stroke="#191970" points="1901.6478,-475.5609 1912.225,-474.9507 1903.5759,-468.8316 1901.6478,-475.5609"/>
</g>
-<!-- Node65 -->
+<!-- Node66 -->
<g id="node13" class="node">
-<title>Node65</title>
+<title>Node66</title>
<g id="a_node13"><a xlink:href="cuda_2reduction_8h.html" target="_top" xlink:title="CUDA schedule for reduction operations. ">
<polygon fill="#ffffff" stroke="#000000" points="1498.5,-335.5 1498.5,-365.5 1620.5,-365.5 1620.5,-335.5 1498.5,-335.5"/>
<text text-anchor="start" x="1506.5" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/cuda</text>
@@ -295,15 +295,15 @@
</a>
</g>
</g>
-<!-- Node56->Node65 -->
+<!-- Node57->Node66 -->
<g id="edge14" class="edge">
-<title>Node56->Node65</title>
+<title>Node57->Node66</title>
<path fill="none" stroke="#191970" d="M1890.0264,-473.1654C1882.1222,-471.7259 1874.1197,-470.3024 1866.5,-469 1763.6464,-451.4204 1726.3965,-482.426 1634.5,-433 1604.8373,-417.0461 1580.8492,-384.6666 1568.5252,-365.5603"/>
<polygon fill="#191970" stroke="#191970" points="1889.4816,-476.6238 1899.9496,-474.9894 1890.7472,-469.7391 1889.4816,-476.6238"/>
</g>
-<!-- Node66 -->
+<!-- Node67 -->
<g id="node14" class="node">
-<title>Node66</title>
+<title>Node67</title>
<g id="a_node14"><a xlink:href="rocm_2reduction_8h.html" target="_top" xlink:title="rocm schedule for reduction operations ">
<polygon fill="#ffffff" stroke="#000000" points="1497.5,-268.5 1497.5,-298.5 1621.5,-298.5 1621.5,-268.5 1497.5,-268.5"/>
<text text-anchor="start" x="1505.5" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/rocm</text>
@@ -311,15 +311,15 @@
</a>
</g>
</g>
-<!-- Node56->Node66 -->
+<!-- Node57->Node67 -->
<g id="edge101" class="edge">
-<title>Node56->Node66</title>
+<title>Node57->Node67</title>
<path fill="none" stroke="#191970" d="M1893.7907,-472.8593C1834.3171,-460.1819 1746.6785,-440.6864 1732.5,-433 1676.95,-402.8853 1677.9339,-375.5814 1629.5,-335 1614.0034,-322.0158 1595.5507,-308.5141 1581.4825,-298.5942"/>
<polygon fill="#191970" stroke="#191970" points="1893.3011,-476.3334 1903.8101,-474.9875 1894.7555,-469.4861 1893.3011,-476.3334"/>
</g>
-<!-- Node67 -->
+<!-- Node68 -->
<g id="node15" class="node">
-<title>Node67</title>
+<title>Node68</title>
<g id="a_node15"><a xlink:href="cuda_2softmax_8h.html" target="_top" xlink:title="include/tvm/topi/cuda\l/softmax.h">
<polygon fill="#ffffff" stroke="#000000" points="189.5,-335.5 189.5,-365.5 311.5,-365.5 311.5,-335.5 189.5,-335.5"/>
<text text-anchor="start" x="197.5" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/cuda</text>
@@ -327,15 +327,15 @@
</a>
</g>
</g>
-<!-- Node56->Node67 -->
+<!-- Node57->Node68 -->
<g id="edge16" class="edge">
-<title>Node56->Node67</title>
+<title>Node57->Node68</title>
<path fill="none" stroke="#191970" d="M1895.8014,-473.1595C1886.0889,-471.4835 1876.0227,-469.9898 1866.5,-469 1778.1756,-459.8192 337.0291,-489.4735 268.5,-433 248.9279,-416.871 247.5885,-384.8653 248.7773,-365.8069"/>
<polygon fill="#191970" stroke="#191970" points="1895.2942,-476.6244 1905.7572,-474.9585 1896.539,-469.736 1895.2942,-476.6244"/>
</g>
-<!-- Node68 -->
+<!-- Node69 -->
<g id="node16" class="node">
-<title>Node68</title>
+<title>Node69</title>
<g id="a_node16"><a xlink:href="rocm_2softmax_8h.html" target="_top" xlink:title="include/tvm/topi/rocm\l/softmax.h">
<polygon fill="#ffffff" stroke="#000000" points="188.5,-268.5 188.5,-298.5 312.5,-298.5 312.5,-268.5 188.5,-268.5"/>
<text text-anchor="start" x="196.5" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/rocm</text>
@@ -343,15 +343,15 @@
</a>
</g>
</g>
-<!-- Node56->Node68 -->
+<!-- Node57->Node69 -->
<g id="edge102" class="edge">
-<title>Node56->Node68</title>
+<title>Node57->Node69</title>
<path fill="none" stroke="#191970" d="M1895.8017,-473.1572C1886.0891,-471.4814 1876.0229,-469.9884 1866.5,-469 1777.0141,-459.7117 330.2305,-468.3506 247.5,-433 198.982,-412.2683 156.6873,-382.0824 180.5,-335 188.4669,-319.2478 203.773,-307.1369 218.0457,-298.6292"/>
<polygon fill="#191970" stroke="#191970" points="1895.2944,-476.6221 1905.7575,-474.9561 1896.5392,-469.7337 1895.2944,-476.6221"/>
</g>
-<!-- Node69 -->
+<!-- Node70 -->
<g id="node17" class="node">
-<title>Node69</title>
+<title>Node70</title>
<g id="a_node17"><a xlink:href="array__utils_8h.html" target="_top" xlink:title="Utility functions for handling arrays. ">
<polygon fill="#ffffff" stroke="#000000" points="1839,-402.5 1839,-432.5 1966,-432.5 1966,-402.5 1839,-402.5"/>
<text text-anchor="start" x="1847" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -359,15 +359,15 @@
</a>
</g>
</g>
-<!-- Node56->Node69 -->
+<!-- Node57->Node70 -->
<g id="edge18" class="edge">
-<title>Node56->Node69</title>
+<title>Node57->Node70</title>
<path fill="none" stroke="#191970" d="M1936.8048,-466.4026C1929.3627,-455.7936 1920.0725,-442.5502 1913.0377,-432.5218"/>
<polygon fill="#191970" stroke="#191970" points="1934.1579,-468.7239 1942.766,-474.9005 1939.8885,-464.7039 1934.1579,-468.7239"/>
</g>
-<!-- Node70 -->
+<!-- Node71 -->
<g id="node18" class="node">
-<title>Node70</title>
+<title>Node71</title>
<g id="a_node18"><a xlink:href="detail_2broadcast_8h.html" target="_top" xlink:title="Detail broadcast. ">
<polygon fill="#ffffff" stroke="#000000" points="2689,-335.5 2689,-365.5 2816,-365.5 2816,-335.5 2689,-335.5"/>
<text text-anchor="start" x="2697" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -375,30 +375,30 @@
</a>
</g>
</g>
-<!-- Node56->Node70 -->
+<!-- Node57->Node71 -->
<g id="edge23" class="edge">
-<title>Node56->Node70</title>
+<title>Node57->Node71</title>
<path fill="none" stroke="#191970" d="M2032.8296,-482.7546C2215.2145,-478.3613 2638.0454,-464.8404 2694.5,-433 2721.1875,-417.9482 2738.4916,-384.9228 2746.7701,-365.5307"/>
<polygon fill="#191970" stroke="#191970" points="2032.6613,-479.2575 2022.7473,-482.9941 2032.8276,-486.2556 2032.6613,-479.2575"/>
</g>
-<!-- Node73 -->
+<!-- Node74 -->
<g id="node21" class="node">
-<title>Node73</title>
+<title>Node74</title>
<g id="a_node21"><a xlink:href="reduction_8h.html" target="_top" xlink:title="Reduction op constructors. ">
<polygon fill="#ffffff" stroke="#000000" points="2862.5,-140 2862.5,-159 3016.5,-159 3016.5,-140 2862.5,-140"/>
<text text-anchor="middle" x="2939.5" y="-147" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/reduction.h</text>
</a>
</g>
</g>
-<!-- Node56->Node73 -->
+<!-- Node57->Node74 -->
<g id="edge97" class="edge">
-<title>Node56->Node73</title>
+<title>Node57->Node74</title>
<path fill="none" stroke="#191970" d="M2033.0288,-483.7296C2284.8075,-481.002 3021.2859,-469.8904 3056.5,-433 3128.645,-357.4207 3085.3681,-281.2857 3018.5,-201 3002.888,-182.2553 2978.6873,-167.8845 2961.1844,-159.146"/>
<polygon fill="#191970" stroke="#191970" points="2032.8034,-480.2317 2022.8413,-483.8383 2032.8781,-487.2313 2032.8034,-480.2317"/>
</g>
-<!-- Node75 -->
+<!-- Node76 -->
<g id="node23" class="node">
-<title>Node75</title>
+<title>Node76</title>
<g id="a_node23"><a xlink:href="nn_2softmax_8h.html" target="_top" xlink:title="Softmax op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="2451,-67.5 2451,-97.5 2562,-97.5 2562,-67.5 2451,-67.5"/>
<text text-anchor="start" x="2459" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -406,21 +406,21 @@
</a>
</g>
</g>
-<!-- Node56->Node75 -->
+<!-- Node57->Node76 -->
<g id="edge95" class="edge">
-<title>Node56->Node75</title>
+<title>Node57->Node76</title>
<path fill="none" stroke="#191970" d="M2032.547,-474.7894C2092.7731,-466.3109 2167.8105,-452.3241 2192.5,-433 2224.589,-407.8844 2232.5,-391.2492 2232.5,-350.5 2232.5,-350.5 2232.5,-350.5 2232.5,-216.5 2232.5,-178.3133 2228.4406,-159.9015 2256.5,-134 2270.7051,-120.8873 2381.7032,-101.7223 2450.938,-90.8485"/>
<polygon fill="#191970" stroke="#191970" points="2032.0665,-471.3224 2022.6376,-476.1541 2033.0216,-478.2569 2032.0665,-471.3224"/>
</g>
-<!-- Node56->Node76 -->
+<!-- Node57->Node77 -->
<g id="edge96" class="edge">
-<title>Node56->Node76</title>
+<title>Node57->Node77</title>
<path fill="none" stroke="#191970" d="M2033.1796,-482.5876C2287.3585,-476.5302 3035.9574,-456.7572 3082.5,-433 3158.1399,-394.3903 3201.5,-368.4241 3201.5,-283.5 3201.5,-283.5 3201.5,-283.5 3201.5,-216.5 3201.5,-165.4534 3163.2148,-113.21 3145.8669,-92.1574"/>
<polygon fill="#191970" stroke="#191970" points="2032.8121,-479.0953 2022.898,-482.8316 2032.9782,-486.0933 2032.8121,-479.0953"/>
</g>
-<!-- Node77 -->
+<!-- Node78 -->
<g id="node25" class="node">
-<title>Node77</title>
+<title>Node78</title>
<g id="a_node25"><a xlink:href="reorg_8h.html" target="_top" xlink:title="Reorg op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="2671.5,-67.5 2671.5,-97.5 2799.5,-97.5 2799.5,-67.5 2671.5,-67.5"/>
<text text-anchor="start" x="2679.5" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/vision</text>
@@ -428,15 +428,15 @@
</a>
</g>
</g>
-<!-- Node56->Node77 -->
+<!-- Node57->Node78 -->
<g id="edge104" class="edge">
-<title>Node56->Node77</title>
+<title>Node57->Node78</title>
<path fill="none" stroke="#191970" d="M2032.9531,-476.2741C2101.3605,-468.1717 2191.5916,-453.9474 2221.5,-433 2256.431,-408.5349 2270.5,-393.1464 2270.5,-350.5 2270.5,-350.5 2270.5,-350.5 2270.5,-216.5 2270.5,-134.0609 2543.7702,-99.2689 2671.3319,-87.5189"/>
<polygon fill="#191970" stroke="#191970" points="2032.1324,-472.8457 2022.6017,-477.4736 2032.9382,-479.7992 2032.1324,-472.8457"/>
</g>
-<!-- Node78 -->
+<!-- Node79 -->
<g id="node26" class="node">
-<title>Node78</title>
+<title>Node79</title>
<g id="a_node26"><a xlink:href="bias__add_8h.html" target="_top" xlink:title="bias_add op constructions ">
<polygon fill="#ffffff" stroke="#000000" points="2666,-134.5 2666,-164.5 2777,-164.5 2777,-134.5 2666,-134.5"/>
<text text-anchor="start" x="2674" y="-152.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -444,57 +444,57 @@
</a>
</g>
</g>
-<!-- Node56->Node78 -->
+<!-- Node57->Node79 -->
<g id="edge86" class="edge">
-<title>Node56->Node78</title>
+<title>Node57->Node79</title>
<path fill="none" stroke="#191970" d="M2032.6125,-481.7373C2119.4394,-477.1498 2248.0141,-465.0083 2284.5,-433 2313.206,-407.8169 2308.5,-388.6867 2308.5,-350.5 2308.5,-350.5 2308.5,-350.5 2308.5,-283.5 2308.5,-209.2497 2553.0697,-169.9234 2665.801,-155.7402"/>
<polygon fill="#191970" stroke="#191970" points="2032.3145,-478.2478 2022.5037,-482.2474 2032.6674,-485.2389 2032.3145,-478.2478"/>
</g>
-<!-- Node79 -->
+<!-- Node80 -->
<g id="node27" class="node">
-<title>Node79</title>
+<title>Node80</title>
<g id="a_node27"><a xlink:href="topi_2transform_8h.html" target="_top" xlink:title="Transform op constructors. ">
<polygon fill="#ffffff" stroke="#000000" points="2537.5,-207 2537.5,-226 2693.5,-226 2693.5,-207 2537.5,-207"/>
<text text-anchor="middle" x="2615.5" y="-214" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/transform.h</text>
</a>
</g>
</g>
-<!-- Node56->Node79 -->
+<!-- Node57->Node80 -->
<g id="edge103" class="edge">
-<title>Node56->Node79</title>
+<title>Node57->Node80</title>
<path fill="none" stroke="#191970" d="M2032.686,-480.8974C2127.5572,-475.4383 2275.6072,-462.3993 2322.5,-433 2397.1735,-386.1837 2361.2092,-319.8221 2432.5,-268 2462.9345,-245.8769 2502.5527,-233.2445 2537.1402,-226.0373"/>
<polygon fill="#191970" stroke="#191970" points="2032.3239,-477.4121 2022.5347,-481.4644 2032.7143,-484.4012 2032.3239,-477.4121"/>
</g>
-<!-- Node56->Node80 -->
+<!-- Node57->Node81 -->
<g id="edge40" class="edge">
-<title>Node56->Node80</title>
+<title>Node57->Node81</title>
<path fill="none" stroke="#191970" d="M2032.8795,-481.3845C2192.163,-475.0682 2550.6531,-459.0961 2851.5,-433 2874.1964,-431.0313 2899.0064,-428.2984 2920.9857,-425.6749"/>
<polygon fill="#191970" stroke="#191970" points="2032.533,-477.8954 2022.6786,-481.7866 2032.8088,-484.89 2032.533,-477.8954"/>
</g>
-<!-- Node82 -->
+<!-- Node83 -->
<g id="node30" class="node">
-<title>Node82</title>
+<title>Node83</title>
<g id="a_node30"><a xlink:href="einsum_8h.html" target="_top" xlink:title="Einstein summation op. ">
<polygon fill="#ffffff" stroke="#000000" points="2412.5,-341 2412.5,-360 2556.5,-360 2556.5,-341 2412.5,-341"/>
<text text-anchor="middle" x="2484.5" y="-348" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/einsum.h</text>
</a>
</g>
</g>
-<!-- Node56->Node82 -->
+<!-- Node57->Node83 -->
<g id="edge82" class="edge">
-<title>Node56->Node82</title>
+<title>Node57->Node83</title>
<path fill="none" stroke="#191970" d="M2033.1681,-482.9075C2181.7273,-479.207 2478.9955,-467.5291 2510.5,-433 2529.6304,-412.033 2506.342,-377.0577 2492.8416,-360.1865"/>
<polygon fill="#191970" stroke="#191970" points="2032.6468,-479.419 2022.7346,-483.1604 2032.8165,-486.417 2032.6468,-479.419"/>
</g>
-<!-- Node56->Node83 -->
+<!-- Node57->Node84 -->
<g id="edge87" class="edge">
-<title>Node56->Node83</title>
+<title>Node57->Node84</title>
<path fill="none" stroke="#191970" d="M2033.0409,-483.6502C2286.5173,-480.6872 3035.8657,-468.9545 3139.5,-433 3161.7582,-425.2778 3162.3429,-414.1965 3182.5,-402 3204.9077,-388.4417 3231.3537,-375.2739 3251.8375,-365.6472"/>
<polygon fill="#191970" stroke="#191970" points="2032.7481,-480.1533 2022.7892,-483.7684 2032.8289,-487.1528 2032.7481,-480.1533"/>
</g>
-<!-- Node84 -->
+<!-- Node85 -->
<g id="node32" class="node">
-<title>Node84</title>
+<title>Node85</title>
<g id="a_node32"><a xlink:href="flatten_8h.html" target="_top" xlink:title="Softmax op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="2834,-335.5 2834,-365.5 2945,-365.5 2945,-335.5 2834,-335.5"/>
<text text-anchor="start" x="2842" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -502,15 +502,15 @@
</a>
</g>
</g>
-<!-- Node56->Node84 -->
+<!-- Node57->Node85 -->
<g id="edge92" class="edge">
-<title>Node56->Node84</title>
+<title>Node57->Node85</title>
<path fill="none" stroke="#191970" d="M2032.7548,-483.1151C2205.2114,-479.5701 2596.1061,-467.844 2724.5,-433 2778.5908,-418.3207 2835.8339,-385.1584 2866.6429,-365.6426"/>
<polygon fill="#191970" stroke="#191970" points="2032.6399,-479.6166 2022.7125,-483.3173 2032.7809,-486.6152 2032.6399,-479.6166"/>
</g>
-<!-- Node85 -->
+<!-- Node86 -->
<g id="node33" class="node">
-<title>Node85</title>
+<title>Node86</title>
<g id="a_node33"><a xlink:href="detail_2extern_8h.html" target="_top" xlink:title="Helpers for using external functions. ">
<polygon fill="#ffffff" stroke="#000000" points="754,-402.5 754,-432.5 881,-432.5 881,-402.5 754,-402.5"/>
<text text-anchor="start" x="762" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -518,15 +518,15 @@
</a>
</g>
</g>
-<!-- Node56->Node85 -->
+<!-- Node57->Node86 -->
<g id="edge52" class="edge">
-<title>Node56->Node85</title>
+<title>Node57->Node86</title>
<path fill="none" stroke="#191970" d="M1894.8234,-473.1315C1885.4117,-471.5071 1875.6996,-470.0379 1866.5,-469 1469.5863,-424.2206 1366.9203,-461.4072 968.5,-433 939.8479,-430.9571 908.2202,-427.8439 881.345,-424.9421"/>
<polygon fill="#191970" stroke="#191970" points="1894.4179,-476.6147 1904.8799,-474.9421 1895.6583,-469.7255 1894.4179,-476.6147"/>
</g>
-<!-- Node86 -->
+<!-- Node87 -->
<g id="node34" class="node">
-<title>Node86</title>
+<title>Node87</title>
<g id="a_node34"><a xlink:href="fuse_8h.html" target="_top" xlink:title="Fuse operation. ">
<polygon fill="#ffffff" stroke="#000000" points="1195,-402.5 1195,-432.5 1322,-432.5 1322,-402.5 1195,-402.5"/>
<text text-anchor="start" x="1203" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -534,15 +534,15 @@
</a>
</g>
</g>
-<!-- Node56->Node86 -->
+<!-- Node57->Node87 -->
<g id="edge55" class="edge">
-<title>Node56->Node86</title>
+<title>Node57->Node87</title>
<path fill="none" stroke="#191970" d="M1893.1788,-473.197C1884.2762,-471.6374 1875.1545,-470.1716 1866.5,-469 1813.6184,-461.8409 1467.303,-434.0454 1322.2451,-422.5351"/>
<polygon fill="#191970" stroke="#191970" points="1892.638,-476.6559 1903.1006,-474.987 1893.8809,-469.7671 1892.638,-476.6559"/>
</g>
-<!-- Node87 -->
+<!-- Node88 -->
<g id="node35" class="node">
-<title>Node87</title>
+<title>Node88</title>
<g id="a_node35"><a xlink:href="generic_2default_8h.html" target="_top" xlink:title="Generic default schedule. ">
<polygon fill="#ffffff" stroke="#000000" points="406,-335.5 406,-365.5 541,-365.5 541,-335.5 406,-335.5"/>
<text text-anchor="start" x="414" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/generic</text>
@@ -550,15 +550,15 @@
</a>
</g>
</g>
-<!-- Node56->Node87 -->
+<!-- Node57->Node88 -->
<g id="edge83" class="edge">
-<title>Node56->Node87</title>
+<title>Node57->Node88</title>
<path fill="none" stroke="#191970" d="M1895.8012,-473.161C1886.0888,-471.4849 1876.0226,-469.9908 1866.5,-469 1822.7249,-464.4454 313.2651,-464.4723 282.5,-433 240.1441,-389.6705 335.3098,-367.7022 405.9555,-357.7052"/>
<polygon fill="#191970" stroke="#191970" points="1895.294,-476.626 1905.757,-474.9602 1896.5389,-469.7376 1895.294,-476.626"/>
</g>
-<!-- Node88 -->
+<!-- Node89 -->
<g id="node36" class="node">
-<title>Node88</title>
+<title>Node89</title>
<g id="a_node36"><a xlink:href="generic_2extern_8h.html" target="_top" xlink:title="Schedule for extern followed by injective ops. ">
<polygon fill="#ffffff" stroke="#000000" points="559,-268.5 559,-298.5 694,-298.5 694,-268.5 559,-268.5"/>
<text text-anchor="start" x="567" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/generic</text>
@@ -566,15 +566,15 @@
</a>
</g>
</g>
-<!-- Node56->Node88 -->
+<!-- Node57->Node89 -->
<g id="edge84" class="edge">
-<title>Node56->Node88</title>
+<title>Node57->Node89</title>
<path fill="none" stroke="#191970" d="M1895.464,-473.161C1885.8547,-471.5018 1875.9108,-470.0135 1866.5,-469 1600.8054,-440.3862 924.9296,-492.9064 664.5,-433 631.7361,-425.4634 627.6119,-411.9548 595.5,-402 509.6505,-375.3863 454.1209,-434.9801 396.5,-366 387.6672,-355.426 387.8875,-345.7542 396.5,-335 399.1098,-331.7412 491.9218,-311.6959 558.8317,-297.5986"/>
<polygon fill="#191970" stroke="#191970" points="1894.8578,-476.608 1905.3204,-474.9391 1896.1006,-469.7192 1894.8578,-476.608"/>
</g>
-<!-- Node89 -->
+<!-- Node90 -->
<g id="node37" class="node">
-<title>Node89</title>
+<title>Node90</title>
<g id="a_node37"><a xlink:href="generic_2injective_8h.html" target="_top" xlink:title="Generic schedule for injective operations. ">
<polygon fill="#ffffff" stroke="#000000" points="559,-335.5 559,-365.5 694,-365.5 694,-335.5 559,-335.5"/>
<text text-anchor="start" x="567" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/generic</text>
@@ -582,15 +582,15 @@
</a>
</g>
</g>
-<!-- Node56->Node89 -->
+<!-- Node57->Node90 -->
<g id="edge85" class="edge">
-<title>Node56->Node89</title>
+<title>Node57->Node90</title>
<path fill="none" stroke="#191970" d="M1895.463,-473.1699C1885.8538,-471.5098 1875.9103,-470.0188 1866.5,-469 1738.5191,-455.1436 826.5723,-484.2839 708.5,-433 676.0854,-418.9209 649.669,-385.5383 636.1975,-365.8229"/>
<polygon fill="#191970" stroke="#191970" points="1894.8567,-476.617 1905.3194,-474.9485 1896.0998,-469.7282 1894.8567,-476.617"/>
</g>
-<!-- Node90 -->
+<!-- Node91 -->
<g id="node38" class="node">
-<title>Node90</title>
+<title>Node91</title>
<g id="a_node38"><a xlink:href="x86_2bnn_8h.html" target="_top" xlink:title="x86 schedule for binary operations ">
<polygon fill="#ffffff" stroke="#000000" points="902,-335.5 902,-365.5 1019,-365.5 1019,-335.5 902,-335.5"/>
<text text-anchor="start" x="910" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/x86</text>
@@ -598,15 +598,15 @@
</a>
</g>
</g>
-<!-- Node56->Node90 -->
+<!-- Node57->Node91 -->
<g id="edge105" class="edge">
-<title>Node56->Node90</title>
+<title>Node57->Node91</title>
<path fill="none" stroke="#191970" d="M1894.4866,-473.1478C1885.1785,-471.5372 1875.5877,-470.0683 1866.5,-469 1544.7644,-431.1792 1453.0337,-510.542 1138.5,-433 1080.4613,-418.6917 1018.3436,-385.2285 985.0163,-365.5981"/>
<polygon fill="#191970" stroke="#191970" points="1893.977,-476.6122 1904.4388,-474.9388 1895.2169,-469.7229 1893.977,-476.6122"/>
</g>
-<!-- Node91 -->
+<!-- Node92 -->
<g id="node39" class="node">
-<title>Node91</title>
+<title>Node92</title>
<g id="a_node39"><a xlink:href="x86_2default_8h.html" target="_top" xlink:title="default x86 schedule ">
<polygon fill="#ffffff" stroke="#000000" points="1037,-335.5 1037,-365.5 1154,-365.5 1154,-335.5 1037,-335.5"/>
<text text-anchor="start" x="1045" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/x86</text>
@@ -614,15 +614,15 @@
</a>
</g>
</g>
-<!-- Node56->Node91 -->
+<!-- Node57->Node92 -->
<g id="edge106" class="edge">
-<title>Node56->Node91</title>
+<title>Node57->Node92</title>
<path fill="none" stroke="#191970" d="M1894.4835,-473.1743C1885.1757,-471.5606 1875.5859,-470.0838 1866.5,-469 1716.022,-451.0513 1325.2359,-491.6485 1185.5,-433 1151.2968,-418.6446 1121.7987,-385.3634 1106.5562,-365.7399"/>
<polygon fill="#191970" stroke="#191970" points="1893.9734,-476.6387 1904.4355,-474.967 1895.2144,-469.7495 1893.9734,-476.6387"/>
</g>
-<!-- Node92 -->
+<!-- Node93 -->
<g id="node40" class="node">
-<title>Node92</title>
+<title>Node93</title>
<g id="a_node40"><a xlink:href="x86_2injective_8h.html" target="_top" xlink:title="x86 schedule for injective ops ">
<polygon fill="#ffffff" stroke="#000000" points="1363,-335.5 1363,-365.5 1480,-365.5 1480,-335.5 1363,-335.5"/>
<text text-anchor="start" x="1371" y="-353.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/x86</text>
@@ -630,15 +630,15 @@
</a>
</g>
</g>
-<!-- Node56->Node92 -->
+<!-- Node57->Node93 -->
<g id="edge107" class="edge">
-<title>Node56->Node92</title>
+<title>Node57->Node93</title>
<path fill="none" stroke="#191970" d="M1890.9565,-473.1623C1882.7556,-471.6923 1874.4248,-470.2616 1866.5,-469 1745.6381,-449.7599 1710.8023,-467.899 1593.5,-433 1538.4253,-416.6145 1478.9989,-384.5448 1446.4166,-365.5766"/>
<polygon fill="#191970" stroke="#191970" points="1890.3939,-476.6173 1900.8592,-474.9659 1891.6483,-469.7306 1890.3939,-476.6173"/>
</g>
-<!-- Node93 -->
+<!-- Node94 -->
<g id="node41" class="node">
-<title>Node93</title>
+<title>Node94</title>
<g id="a_node41"><a xlink:href="pad__utils_8h.html" target="_top" xlink:title="Padding helpers. ">
<polygon fill="#ffffff" stroke="#000000" points="2077,-201.5 2077,-231.5 2204,-231.5 2204,-201.5 2077,-201.5"/>
<text text-anchor="start" x="2085" y="-219.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -646,15 +646,15 @@
</a>
</g>
</g>
-<!-- Node56->Node93 -->
+<!-- Node57->Node94 -->
<g id="edge73" class="edge">
-<title>Node56->Node93</title>
+<title>Node57->Node94</title>
<path fill="none" stroke="#191970" d="M2032.7989,-475.8251C2087.9663,-468.1631 2153.2355,-454.7503 2170.5,-433 2218.8624,-372.0719 2170.5319,-269.8932 2149.3558,-231.5995"/>
<polygon fill="#191970" stroke="#191970" points="2032.0751,-472.3908 2022.6294,-477.1896 2033.006,-479.3286 2032.0751,-472.3908"/>
</g>
-<!-- Node94 -->
+<!-- Node95 -->
<g id="node42" class="node">
-<title>Node94</title>
+<title>Node95</title>
<g id="a_node42"><a xlink:href="ravel__unravel_8h.html" target="_top" xlink:title="Index ravel and unraval operations. ">
<polygon fill="#ffffff" stroke="#000000" points="2559,-402.5 2559,-432.5 2686,-432.5 2686,-402.5 2559,-402.5"/>
<text text-anchor="start" x="2567" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -662,15 +662,15 @@
</a>
</g>
</g>
-<!-- Node56->Node94 -->
+<!-- Node57->Node95 -->
<g id="edge75" class="edge">
-<title>Node56->Node94</title>
+<title>Node57->Node95</title>
<path fill="none" stroke="#191970" d="M2032.7811,-480.1851C2148.1421,-473.5503 2362.8973,-458.8235 2544.5,-433 2549.1474,-432.3391 2553.9363,-431.5804 2558.7469,-430.7615"/>
<polygon fill="#191970" stroke="#191970" points="2032.3962,-476.7012 2022.6112,-480.7638 2032.7939,-483.6899 2032.3962,-476.7012"/>
</g>
-<!-- Node95 -->
+<!-- Node96 -->
<g id="node43" class="node">
-<title>Node95</title>
+<title>Node96</title>
<g id="a_node43"><a xlink:href="tensor__utils_8h.html" target="_top" xlink:title="Utility functions for handling tensor. ">
<polygon fill="#ffffff" stroke="#000000" points="2375,-402.5 2375,-432.5 2502,-432.5 2502,-402.5 2375,-402.5"/>
<text text-anchor="start" x="2383" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/detail</text>
@@ -678,15 +678,15 @@
</a>
</g>
</g>
-<!-- Node56->Node95 -->
+<!-- Node57->Node96 -->
<g id="edge79" class="edge">
-<title>Node56->Node95</title>
+<title>Node57->Node96</title>
<path fill="none" stroke="#191970" d="M2033.0118,-475.8069C2116.2302,-466.7282 2247.547,-451.2756 2360.5,-433 2365.1339,-432.2502 2369.9128,-431.4262 2374.7164,-430.5609"/>
<polygon fill="#191970" stroke="#191970" points="2032.4187,-472.3507 2022.8548,-476.9094 2033.1741,-479.3098 2032.4187,-472.3507"/>
</g>
-<!-- Node96 -->
+<!-- Node97 -->
<g id="node44" class="node">
-<title>Node96</title>
+<title>Node97</title>
<g id="a_node44"><a xlink:href="nn_2dense_8h.html" target="_top" xlink:title="Dense op constructions. ">
<polygon fill="#ffffff" stroke="#000000" points="430,-268.5 430,-298.5 541,-298.5 541,-268.5 430,-268.5"/>
<text text-anchor="start" x="438" y="-286.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/topi/nn</text>
@@ -694,21 +694,21 @@
</a>
</g>
</g>
-<!-- Node56->Node96 -->
+<!-- Node57->Node97 -->
<g id="edge88" class="edge">
... 278252 lines suppressed ...