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 2022/05/10 07:06:07 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@8d4f4dd73ff2d762ade38aa0f9e8e9fba5f933da)
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 f81d4f7ee deploying docs (apache/tvm@8d4f4dd73ff2d762ade38aa0f9e8e9fba5f933da)
f81d4f7ee is described below
commit f81d4f7ee8157ac94ea2f5624b27515757ac28c3
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Tue May 10 07:06:00 2022 +0000
deploying docs (apache/tvm@8d4f4dd73ff2d762ade38aa0f9e8e9fba5f933da)
---
.../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 +-
.../how_to/compile_models/from_tensorflow.rst.txt | 5 +
.../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 | 1172 +---
.../tune_network_cuda.rst.txt | 2 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 87 +-
.../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 +-
.../how_to/work_with_microtvm/micro_tflite.rst.txt | 1 +
.../work_with_microtvm/sg_execution_times.rst.txt | 12 +-
.../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 | 10 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 62 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 26 +-
.../tutorial/tensor_expr_get_started.rst.txt | 45 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 92 +-
docs/how_to/compile_models/from_paddle.html | 2 +-
docs/how_to/compile_models/from_pytorch.html | 8 +-
docs/how_to/compile_models/from_tensorflow.html | 1 +
docs/how_to/compile_models/sg_execution_times.html | 22 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 72 +-
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 | 1171 +---
.../tune_with_autoscheduler/tune_network_cuda.html | 2 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 86 +-
.../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 +-
docs/how_to/work_with_microtvm/micro_tflite.html | 1 +
.../work_with_microtvm/sg_execution_times.html | 12 +-
.../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/annotated.html | 43 +-
docs/reference/api/doxygen/aot__executor_8h.html | 370 ++
.../api/doxygen/aot__executor_8h__incl.svg | 119 +
.../api/doxygen/aot__executor_8h_source.html | 93 +
...odule_8h.html => aot__executor__module_8h.html} | 24 +-
.../api/doxygen/aot__executor__module_8h__incl.svg | 36 +
.../doxygen/aot__executor__module_8h_source.html | 80 +
docs/reference/api/doxygen/array_8h__dep__incl.svg | 204 +-
.../api/doxygen/c__backend__api_8h__dep__incl.svg | 2 +-
docs/reference/api/doxygen/c__runtime__api_8h.html | 6 +-
.../api/doxygen/c__runtime__api_8h__dep__incl.svg | 1296 ++---
docs/reference/api/doxygen/classes.html | 429 +-
...asstvm_1_1runtime_1_1ObjectRef__coll__graph.svg | 12 +-
.../classtvm_1_1runtime_1_1Object__coll__graph.svg | 8 +-
...vm_1_1runtime_1_1metadata_1_1ArrayAccessor.html | 4 +-
...1_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html | 4 +-
...vm_1_1runtime_1_1metadata_1_1ArrayIterator.html | 4 +-
...1_1runtime_1_1metadata_1_1Metadata-members.html | 102 +
...asstvm_1_1runtime_1_1metadata_1_1Metadata.html} | 75 +-
..._1runtime_1_1metadata_1_1MetadataArrayNode.html | 4 +-
...tvm_1_1runtime_1_1metadata_1_1MetadataBase.html | 2 +-
...1_1runtime_1_1metadata_1_1MetadataBaseNode.html | 10 +-
..._1metadata_1_1MetadataBaseNode__coll__graph.svg | 16 +-
...etadata_1_1MetadataBaseNode__inherit__graph.svg | 190 +-
...me_1_1metadata_1_1MetadataBase__coll__graph.svg | 20 +-
...1_1metadata_1_1MetadataBase__inherit__graph.svg | 124 +-
...untime_1_1metadata_1_1MetadataNode-members.html | 124 +
...vm_1_1runtime_1_1metadata_1_1MetadataNode.html} | 304 +-
...me_1_1metadata_1_1MetadataNode__coll__graph.svg | 112 +
...1_1metadata_1_1MetadataNode__inherit__graph.svg | 106 +
...untime_1_1metadata_1_1Metadata__coll__graph.svg | 112 +
...me_1_1metadata_1_1Metadata__inherit__graph.svg} | 54 +-
...1runtime_1_1metadata_1_1TensorInfo-members.html | 102 +
...stvm_1_1runtime_1_1metadata_1_1TensorInfo.html} | 75 +-
...time_1_1metadata_1_1TensorInfoNode-members.html | 120 +
..._1_1runtime_1_1metadata_1_1TensorInfoNode.html} | 232 +-
..._1_1metadata_1_1TensorInfoNode__coll__graph.svg | 108 +
...1metadata_1_1TensorInfoNode__inherit__graph.svg | 102 +
...time_1_1metadata_1_1TensorInfo__coll__graph.svg | 112 +
..._1_1metadata_1_1TensorInfo__inherit__graph.svg} | 54 +-
.../doxygen/crt_2packed__func_8h__dep__incl.svg | 2 +-
.../api/doxygen/data__type_8h__dep__incl.svg | 440 +-
...df60bd528724ee5.html => dir_000008_000013.html} | 23 +-
.../dir_02be2c9d68e402f80df60bd528724ee5.html | 2 +-
.../dir_02be2c9d68e402f80df60bd528724ee5_dep.svg | 212 +-
.../dir_404558507ed35459f0d7a6d81d8c508d.html | 3 +
.../dir_404558507ed35459f0d7a6d81d8c508d_dep.svg | 47 +-
.../dir_4378f18824ae7d4ad48f8d7785cd7ac8.html | 2 +-
.../dir_4378f18824ae7d4ad48f8d7785cd7ac8_dep.svg | 226 +-
.../dir_5da96592f3a7c442b838b075c58254c2.html | 2 +-
.../dir_5da96592f3a7c442b838b075c58254c2_dep.svg | 176 +-
.../dir_72c2f11201cd7636dc7624de0754daa5.html | 2 +-
.../dir_72c2f11201cd7636dc7624de0754daa5_dep.svg | 118 +-
.../dir_a2900df4deca8dd2bcded616f0fe650a.html | 8 +-
.../dir_b4c7d8e826c599ba55146c099a14beb5.html | 2 +-
.../dir_b4c7d8e826c599ba55146c099a14beb5_dep.svg | 510 +-
.../dir_dc867ff9a37cad1764f1670dc7eba6c1.html | 2 +-
.../dir_dc867ff9a37cad1764f1670dc7eba6c1_dep.svg | 194 +-
.../dir_f97d855a3173728370e632aa77170e34_dep.svg | 32 +-
.../dir_fb1b1bc11a768ab8cf63a96a73170118_dep.svg | 20 +-
docs/reference/api/doxygen/error__codes_8h.html | 20 +-
.../api/doxygen/error__codes_8h__dep__incl.svg | 256 +-
.../api/doxygen/error__codes_8h_source.html | 10 +-
docs/reference/api/doxygen/files.html | 41 +-
.../api/doxygen/func__registry_8h__dep__incl.svg | 2 +-
docs/reference/api/doxygen/functions__.html | 2 +
docs/reference/api/doxygen/functions_a.html | 7 +-
docs/reference/api/doxygen/functions_d.html | 6 +-
docs/reference/api/doxygen/functions_func_d.html | 9 +-
docs/reference/api/doxygen/functions_func_g.html | 6 +-
docs/reference/api/doxygen/functions_func_i.html | 7 +-
docs/reference/api/doxygen/functions_func_m.html | 9 +
docs/reference/api/doxygen/functions_func_n.html | 15 +-
docs/reference/api/doxygen/functions_func_o.html | 21 +-
docs/reference/api/doxygen/functions_func_p.html | 3 +
docs/reference/api/doxygen/functions_func_s.html | 17 +-
docs/reference/api/doxygen/functions_func_t.html | 12 +-
docs/reference/api/doxygen/functions_func_v.html | 31 +-
docs/reference/api/doxygen/functions_g.html | 2 +
docs/reference/api/doxygen/functions_i.html | 13 +-
docs/reference/api/doxygen/functions_m.html | 19 +-
docs/reference/api/doxygen/functions_n.html | 16 +-
docs/reference/api/doxygen/functions_o.html | 13 +-
docs/reference/api/doxygen/functions_p.html | 9 +-
docs/reference/api/doxygen/functions_s.html | 11 +-
docs/reference/api/doxygen/functions_t.html | 14 +-
docs/reference/api/doxygen/functions_v.html | 39 +-
docs/reference/api/doxygen/functions_vars.html | 2 +
docs/reference/api/doxygen/functions_vars_a.html | 1 +
docs/reference/api/doxygen/functions_vars_d.html | 1 +
docs/reference/api/doxygen/functions_vars_m.html | 4 +
docs/reference/api/doxygen/functions_vars_n.html | 1 +
.../api/doxygen/functor_8h__dep__incl.svg | 272 +-
docs/reference/api/doxygen/globals_eval.html | 24 +-
docs/reference/api/doxygen/globals_func.html | 21 +
docs/reference/api/doxygen/globals_k.html | 24 +-
docs/reference/api/doxygen/globals_t.html | 27 +
docs/reference/api/doxygen/globals_type.html | 6 +
docs/reference/api/doxygen/graph__executor_8h.html | 5 +-
.../api/doxygen/graph__executor__module_8h.html | 7 +-
docs/reference/api/doxygen/hierarchy.html | 183 +-
docs/reference/api/doxygen/inherit_graph_10.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_107.svg | 5694 ++++++++++----------
docs/reference/api/doxygen/inherit_graph_116.svg | 1406 ++---
docs/reference/api/doxygen/inherit_graph_185.svg | 8 +-
docs/reference/api/doxygen/inherit_graph_199.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_200.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_211.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_212.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_213.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_214.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_215.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_216.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_217.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_218.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_219.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_220.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_221.svg | 12 +-
docs/reference/api/doxygen/inherit_graph_222.svg | 12 +-
...inherit_graph_222.svg => inherit_graph_223.svg} | 0
docs/reference/api/doxygen/inherit_graph_39.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_42.svg | 8 +-
docs/reference/api/doxygen/inherit_graph_43.svg | 8 +-
docs/reference/api/doxygen/inherits.html | 28 +-
.../api/doxygen/ir_2expr_8h__dep__incl.svg | 118 +-
.../api/doxygen/ir_2span_8h__dep__incl.svg | 218 +-
.../api/doxygen/ir_2type_8h__dep__incl.svg | 118 +-
docs/reference/api/doxygen/map_8h__dep__incl.svg | 212 +-
docs/reference/api/doxygen/metadata_8h.html | 33 +-
docs/reference/api/doxygen/metadata_8h__incl.svg | 1407 ++++-
docs/reference/api/doxygen/metadata_8h_source.html | 39 +-
docs/reference/api/doxygen/metadata__base_8h.html | 16 +-
.../api/doxygen/metadata__base_8h__dep__incl.svg | 36 +
.../api/doxygen/metadata__base_8h_source.html | 2 +-
.../{metadata_8h.html => metadata__types_8h.html} | 45 +-
.../api/doxygen/metadata__types_8h__dep__incl.svg | 52 +
...a_8h__incl.svg => metadata__types_8h__incl.svg} | 6 +-
.../api/doxygen/metadata__types_8h_source.html | 92 +
.../api/doxygen/namespacemembers_func_p.html | 6 +-
docs/reference/api/doxygen/namespacemembers_p.html | 6 +-
.../namespacetvm_1_1runtime_1_1metadata.html | 18 +-
.../api/doxygen/ndarray_8h__dep__incl.svg | 248 +-
docs/reference/api/doxygen/node_8h__dep__incl.svg | 230 +-
docs/reference/api/doxygen/object_8h.html | 2 +-
.../reference/api/doxygen/object_8h__dep__incl.svg | 1517 +++---
.../api/doxygen/optional_8h__dep__incl.svg | 252 +-
.../api/doxygen/packed__func_8h__dep__incl.svg | 156 +-
.../api/doxygen/platform_8h__dep__incl.svg | 2 +-
.../api/doxygen/reflection_8h__dep__incl.svg | 164 +-
.../api/doxygen/repr__printer_8h__dep__incl.svg | 226 +-
.../runtime_2container_2base_8h__dep__incl.svg | 296 +-
.../doxygen/runtime_2crt_2module_8h__dep__incl.svg | 2 +-
.../api/doxygen/runtime_2memory_8h__dep__incl.svg | 456 +-
.../api/doxygen/runtime_2module_8h__dep__incl.svg | 136 +-
docs/reference/api/doxygen/search/all_1.js | 2 +-
docs/reference/api/doxygen/search/all_10.js | 4 +-
docs/reference/api/doxygen/search/all_11.js | 4 +-
docs/reference/api/doxygen/search/all_13.js | 6 +-
docs/reference/api/doxygen/search/all_14.js | 8 +-
docs/reference/api/doxygen/search/all_15.js | 24 +-
docs/reference/api/doxygen/search/all_16.js | 4 +-
docs/reference/api/doxygen/search/all_17.js | 6 +-
docs/reference/api/doxygen/search/all_2.js | 6 +-
docs/reference/api/doxygen/search/all_5.js | 6 +-
docs/reference/api/doxygen/search/all_7.js | 2 +-
docs/reference/api/doxygen/search/all_8.js | 2 +-
docs/reference/api/doxygen/search/all_9.js | 2 +-
docs/reference/api/doxygen/search/all_a.js | 2 +-
docs/reference/api/doxygen/search/all_c.js | 8 +-
docs/reference/api/doxygen/search/all_e.js | 9 +-
docs/reference/api/doxygen/search/all_f.js | 12 +-
docs/reference/api/doxygen/search/classes_10.js | 2 +-
docs/reference/api/doxygen/search/classes_11.js | 7 +-
docs/reference/api/doxygen/search/classes_13.js | 2 +-
docs/reference/api/doxygen/search/classes_5.js | 2 +-
docs/reference/api/doxygen/search/classes_7.js | 2 +-
docs/reference/api/doxygen/search/classes_a.js | 2 +
docs/reference/api/doxygen/search/enumvalues_5.js | 8 +-
docs/reference/api/doxygen/search/files_0.js | 2 +
docs/reference/api/doxygen/search/files_9.js | 1 +
docs/reference/api/doxygen/search/functions_10.js | 3 +-
docs/reference/api/doxygen/search/functions_12.js | 2 +-
docs/reference/api/doxygen/search/functions_13.js | 4 +-
docs/reference/api/doxygen/search/functions_14.js | 16 +-
docs/reference/api/doxygen/search/functions_15.js | 2 +-
docs/reference/api/doxygen/search/functions_16.js | 3 +-
docs/reference/api/doxygen/search/functions_4.js | 4 +-
docs/reference/api/doxygen/search/functions_7.js | 2 +-
docs/reference/api/doxygen/search/functions_9.js | 1 +
docs/reference/api/doxygen/search/functions_d.js | 3 +
docs/reference/api/doxygen/search/functions_e.js | 7 +-
docs/reference/api/doxygen/search/functions_f.js | 1 +
docs/reference/api/doxygen/search/typedefs_e.js | 2 +
docs/reference/api/doxygen/search/variables_0.js | 2 +-
docs/reference/api/doxygen/search/variables_1.js | 2 +-
docs/reference/api/doxygen/search/variables_4.js | 2 +-
docs/reference/api/doxygen/search/variables_c.js | 3 +-
docs/reference/api/doxygen/search/variables_d.js | 2 +-
.../api/doxygen/serializer_8h__dep__incl.svg | 236 +-
.../api/doxygen/shape__tuple_8h__dep__incl.svg | 220 +-
.../reference/api/doxygen/string_8h__dep__incl.svg | 174 +-
...4ee5.html => structTVMAotExecutor-members.html} | 31 +-
.../api/doxygen/structTVMAotExecutor.html | 184 +
.../doxygen/structTVMAotExecutor__coll__graph.svg | 75 +
docs/reference/api/doxygen/structTVMMetadata.html | 4 +-
.../reference/api/doxygen/structTVMTensorInfo.html | 4 +-
.../doxygen/structural__equal_8h__dep__incl.svg | 196 +-
.../api/doxygen/structural__hash_8h__dep__incl.svg | 196 +-
docs/reference/api/doxygen/support_2span_8h.html | 6 +
.../api/doxygen/support_2span_8h__dep__incl.svg | 36 +
docs/reference/api/python/auto_scheduler.html | 6 +-
.../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 | 6 +-
docs/tutorial/autotvm_relay_x86.html | 121 +-
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 26 +-
docs/tutorial/tensor_expr_get_started.html | 41 +-
327 files changed, 15144 insertions(+), 12173 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 46362bee5..e86e87983 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.zipf92e990a-8356-4320-b33c-7d175eacbac8 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipdd6354a5-a1b5-4986-b1e1-1c339d1b095f 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 fe0961266..b8696d27d 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<07:31, 96.2kB/s]
0%| | 40.0k/41.5M [00:00<05:49, 124kB/s]
0%| | 88.0k/41.5M [00:00<03:38, 199kB/s]
0%| | 160k/41.5M [00:00<02:29, 290kB/s]
1%| | 336k/41.5M [00:00<01:16, 566kB/s]
1%|1 | 496k/41.5M [00:01<01:01, 699kB/s]
2%|2 | 0.98M/41.5M [00:01<00:29, 1.46MB/s]
4%|4 | 1.70M/41.5M [00:01<00:17, 2.40MB/s]
7%|6 | 2.85M/41.5M [00:01<00:13, 2.94MB/s]
10%|# | 4.32M/41.5M [00:01<00:08, 4.47MB/s]
13%|#3 | 5.48M/41.5M [00:02<00:07, 5.14MB/s]
14%|#4 | 5.99M/41.5M [00:02<00:08, 4.59MB/s]
16%|#6 | 6.66M/41.5M [00:02<00:08, 4.49MB/s]
19%|#9 | 7.95M/41.5M [00:02<00:08, 4.24MB/s]
23%|##2 | 9.42M/41.5M [00:02<00:06, 5.37MB/s]
25%|##5 | 10.4M/41.5M [00:03<00:05, 5.57MB/s]
28%|##7 | 11.4M/41.5M [00:03<00
:05, 5.73MB/s]
30%|##9 | 12.4M/41.5M [00:03<00:05, 5.88MB/s]
32%|###2 | 13.5M/41.5M [00:03<00:04, 6.01MB/s]
35%|###5 | 14.5M/41.5M [00:03<00:04, 6.13MB/s]
38%|###7 | 15.6M/41.5M [00:03<00:04, 6.24MB/s]
40%|#### | 16.7M/41.5M [00:04<00:04, 6.33MB/s]
43%|####2 | 17.7M/41.5M [00:04<00:03, 6.41MB/s]
45%|####5 | 18.8M/41.5M [00:04<00:03, 6.95MB/s]
48%|####7 | 19.9M/41.5M [00:04<00:03, 7.44MB/s]
50%|##### | 20.9M/41.5M [00:04<00:02, 7.55MB/s]
52%|#####2 | 21.6M/41.5M [00:04<00:02, 7.08MB/s]
54%|#####3 | 22.3M/41.5M [00:04<00:03, 6.14MB/s]
56%|#####6 | 23.3M/41.5M [00:05<00:02, 6.66MB/s]
59%|#####8 | 24.4M/41.5M [00:05<00:02, 7.31MB/s]
61%|######1 | 25.4M/41.5M [00:05<00:02, 7.55MB/s]
63%|######2 | 26.1M/41.5M [00:05<00:02, 7.06MB/s]
65%|######4 | 26.8M/41.5M [00:05<00:02, 6.11MB/s]
67%|######7 | 27.9M/41.5M [00:05<00:02, 6.78MB/s]
70%|####
##9 | 29.0M/41.5M [00:05<00:01, 7.48MB/s]
72%|#######2 | 30.0M/41.5M [00:06<00:01, 8.16MB/s]
74%|#######4 | 30.8M/41.5M [00:06<00:01, 7.21MB/s]
76%|#######5 | 31.5M/41.5M [00:06<00:01, 6.28MB/s]
78%|#######8 | 32.5M/41.5M [00:06<00:01, 6.80MB/s]
81%|########1 | 33.7M/41.5M [00:06<00:01, 7.54MB/s]
84%|########3 | 34.7M/41.5M [00:06<00:00, 8.19MB/s]
86%|########5 | 35.5M/41.5M [00:06<00:00, 7.25MB/s]
87%|########7 | 36.2M/41.5M [00:07<00:00, 6.33MB/s]
90%|########9 | 37.3M/41.5M [00:07<00:00, 6.89MB/s]
93%|#########2| 38.4M/41.5M [00:07<00:00, 7.65MB/s]
95%|#########5| 39.5M/41.5M [00:07<00:00, 8.32MB/s]
97%|#########7| 40.3M/41.5M [00:07<00:00, 7.37MB/s]
99%|#########8| 41.0M/41.5M [00:07<00:00, 6.44MB/s]
100%|##########| 41.5M/41.5M [00:07<00:00, 5.64MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
0%| | 16.0k/41.5M [00:00<07:42, 94.0kB/s]
0%| | 48.0k/41.5M [00:00<04:52, 149kB/s]
0%| | 96.0k/41.5M [00:00<03:27, 209kB/s]
0%| | 168k/41.5M [00:00<02:28, 292kB/s]
1%| | 344k/41.5M [00:00<01:17, 558kB/s]
1%|1 | 560k/41.5M [00:01<00:53, 797kB/s]
3%|2 | 1.09M/41.5M [00:01<00:26, 1.61MB/s]
5%|4 | 1.99M/41.5M [00:01<00:14, 2.79MB/s]
7%|7 | 3.00M/41.5M [00:01<00:10, 3.79MB/s]
10%|# | 4.34M/41.5M [00:01<00:07, 5.13MB/s]
14%|#4 | 5.81M/41.5M [00:01<00:05, 6.25MB/s]
18%|#7 | 7.28M/41.5M [00:02<00:05, 7.02MB/s]
21%|##1 | 8.74M/41.5M [00:02<00:03, 8.66MB/s]
24%|##3 | 9.81M/41.5M [00:02<00:03, 9.23MB/s]
26%|##5 | 10.8M/41.5M [00:02<00:03, 8.28MB/s]
28%|##8 | 11.7M/41.5M [00:02<00:03, 7.92MB/s]
32%|###1 | 13.1M/41.5M [00:02<00
:03, 9.36MB/s]
34%|###3 | 14.1M/41.5M [00:02<00:03, 9.08MB/s]
36%|###6 | 15.0M/41.5M [00:02<00:03, 8.02MB/s]
39%|###8 | 16.1M/41.5M [00:03<00:03, 8.10MB/s]
42%|####2 | 17.5M/41.5M [00:03<00:02, 9.14MB/s]
44%|####4 | 18.5M/41.5M [00:03<00:02, 9.26MB/s]
47%|####6 | 19.4M/41.5M [00:03<00:02, 8.14MB/s]
49%|####9 | 20.5M/41.5M [00:03<00:02, 7.71MB/s]
53%|#####2 | 21.9M/41.5M [00:03<00:02, 8.61MB/s]
56%|#####6 | 23.4M/41.5M [00:03<00:02, 9.44MB/s]
59%|#####8 | 24.3M/41.5M [00:04<00:01, 9.49MB/s]
61%|###### | 25.2M/41.5M [00:04<00:02, 8.33MB/s]
63%|######3 | 26.3M/41.5M [00:04<00:01, 8.36MB/s]
67%|######6 | 27.8M/41.5M [00:04<00:01, 9.31MB/s]
69%|######9 | 28.7M/41.5M [00:04<00:01, 9.35MB/s]
71%|#######1 | 29.6M/41.5M [00:04<00:01, 8.18MB/s]
74%|#######4 | 30.7M/41.5M [00:04<00:01, 8.33MB/s]
78%|#######7 | 32.2M/41.5M [00:05<00:01, 8.53MB/s]
81%|####
####1 | 33.7M/41.5M [00:05<00:00, 8.57MB/s]
85%|########4 | 35.1M/41.5M [00:05<00:00, 9.86MB/s]
87%|########7 | 36.1M/41.5M [00:05<00:00, 9.41MB/s]
89%|########9 | 37.0M/41.5M [00:05<00:00, 8.34MB/s]
92%|#########1| 38.1M/41.5M [00:05<00:00, 8.53MB/s]
95%|#########5| 39.5M/41.5M [00:05<00:00, 9.54MB/s]
98%|#########7| 40.5M/41.5M [00:05<00:00, 9.11MB/s]
100%|#########9| 41.3M/41.5M [00:06<00:00, 8.01MB/s]
100%|##########| 41.5M/41.5M [00:06<00:00, 7.08MB/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 100c22d00..9e05dda97 100644
--- a/docs/_sources/how_to/compile_models/from_paddle.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_paddle.rst.txt
@@ -201,7 +201,7 @@ Look up prediction top 1 index in 1000 class synset.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 13.489 seconds)
+ **Total running time of the script:** ( 1 minutes 3.912 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 2cfc6908c..7744ef6d6 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]
40%|###9 | 17.7M/44.7M [00:00<00:00, 185MB/s]
94%|#########3| 41.8M/44.7M [00:00<00:00, 225MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 222MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
5%|5 | 2.35M/44.7M [00:00<00:01, 24.3MB/s]
12%|#1 | 5.17M/44.7M [00:00<00:01, 27.4MB/s]
48%|####7 | 21.4M/44.7M [00:00<00:00, 92.2MB/s]
82%|########2 | 36.7M/44.7M [00:00<00:00, 118MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 99.4MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
index 43d9b260b..7ad506abe 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -370,6 +370,11 @@ Run the corresponding model on tensorflow
+.. rst-class:: sphx-glr-timing
+
+ **Total running time of the script:** ( 1 minutes 1.307 seconds)
+
+
.. _sphx_glr_download_how_to_compile_models_from_tensorflow.py:
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 e3d5e8572..ff8bb1ed6 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
=================
-**05:16.343** total execution time for **how_to_compile_models** files:
+**05:10.309** total execution time for **how_to_compile_models** files:
-- **01:13.489**: :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)
-- **00:58.690**: :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``)
-- **00:54.078**: :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)
-- **00:30.502**: :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)
-- **00:25.434**: :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)
-- **00:21.068**: :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)
-- **00:20.620**: :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)
-- **00:17.985**: :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)
-- **00:12.122**: :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)
-- **00:02.355**: :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)
+- **01:03.912**: :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)
+- **01:01.307**: :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``)
+- **00:55.356**: :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)
+- **00:29.339**: :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)
+- **00:24.634**: :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)
+- **00:21.462**: :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)
+- **00:20.545**: :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)
+- **00:19.195**: :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)
+- **00:11.792**: :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)
+- **00:02.766**: :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 1ca77f9da..60e2606e5 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
@@ -393,7 +393,7 @@ Execute on TVM
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.3286 15.3173 15.4534 15.2190 0.0801
+ 15.5724 15.5598 15.6589 15.5272 0.0414
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 99c8b06cb..4071fb129 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]
10%|9 | 16.5M/170M [00:00<00:00, 174MB/s]
22%|##1 | 37.3M/170M [00:00<00:00, 199MB/s]
36%|###5 | 60.7M/170M [00:00<00:00, 220MB/s]
50%|####9 | 84.2M/170M [00:00<00:00, 231MB/s]
64%|######3 | 108M/170M [00:00<00:00, 238MB/s]
77%|#######7 | 131M/170M [00:00<00:00, 225MB/s]
90%|########9 | 153M/170M [00:00<00:00, 226MB/s]
100%|##########| 170M/170M [00:00<00:00, 225MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
1%|1 | 1.88M/170M [00:00<00:11, 15.7MB/s]
3%|3 | 5.31M/170M [00:00<00:07, 24.2MB/s]
5%|4 | 8.24M/170M [00:00<00:06, 26.9MB/s]
8%|7 | 13.0M/170M [00:00<00:04, 35.4MB/s]
10%|# | 17.4M/170M [00:00<00:04, 39.0MB/s]
12%|#2 | 21.2M/170M [00:00<00:04, 38.2MB/s]
15%|#4 | 24.9M/170M [00:00<00:04, 37.4MB/s]
17%|#6 | 28.5M/170M [00:00<00:05, 28.1MB/s]
19%|#8 | 31.5M/170M [00:01<00:05, 25.7MB/s]
20%|## | 34.1M/170M [00:01<00:07, 18.0MB/s]
21%|##1 | 36.3M/170M [00:01<00:07, 18.3MB/s]
23%|##2 | 38.5M/170M [00:01<00:07, 19.3MB/s]
24%|##3 | 40.6M/170M [00:01<00:07, 18.6MB/s]
25%|##5 | 42.6M/170M [00:01<00:07, 19.0MB/s]
27%|##6 | 45.8M/170M [00:01<00:05, 22.7MB/s]
28%|##8 | 48.1M/170M [00:02<00:05, 21.9MB/s]
30%|##9 | 50.3M/170M [00:02<00:05, 21.9MB/
s]
31%|###1 | 53.1M/170M [00:02<00:05, 23.8MB/s]
33%|###2 | 55.5M/170M [00:02<00:05, 23.8MB/s]
34%|###4 | 57.8M/170M [00:02<00:05, 20.9MB/s]
35%|###5 | 60.2M/170M [00:02<00:05, 22.1MB/s]
38%|###7 | 63.8M/170M [00:02<00:04, 23.2MB/s]
40%|#### | 68.6M/170M [00:02<00:03, 28.4MB/s]
43%|####2 | 72.3M/170M [00:03<00:03, 28.6MB/s]
44%|####4 | 75.0M/170M [00:03<00:03, 25.3MB/s]
46%|####5 | 77.7M/170M [00:03<00:03, 25.9MB/s]
48%|####7 | 80.8M/170M [00:03<00:03, 27.6MB/s]
49%|####9 | 83.5M/170M [00:03<00:03, 25.2MB/s]
51%|##### | 85.9M/170M [00:03<00:04, 21.8MB/s]
52%|#####1 | 88.1M/170M [00:03<00:04, 18.5MB/s]
53%|#####3 | 90.0M/170M [00:04<00:04, 18.2MB/s]
54%|#####4 | 91.8M/170M [00:04<00:04, 18.3MB/s]
55%|#####5 | 93.8M/170M [00:04<00:04, 18.7MB/s]
56%|#####6 | 95.6M/170M [00:04<00:04, 17.3MB/s]
57%|#####7 | 97.4M/170M [00:04<00:
04, 17.6MB/s]
58%|#####8 | 99.1M/170M [00:04<00:04, 17.8MB/s]
59%|#####9 | 101M/170M [00:04<00:04, 17.2MB/s]
60%|###### | 103M/170M [00:04<00:04, 16.1MB/s]
61%|######1 | 104M/170M [00:04<00:04, 14.5MB/s]
62%|######2 | 106M/170M [00:05<00:04, 15.7MB/s]
64%|######3 | 108M/170M [00:05<00:03, 16.9MB/s]
65%|######4 | 110M/170M [00:05<00:03, 19.4MB/s]
67%|######6 | 114M/170M [00:05<00:02, 23.1MB/s]
69%|######9 | 118M/170M [00:05<00:01, 29.2MB/s]
71%|#######1 | 121M/170M [00:05<00:01, 29.4MB/s]
73%|#######3 | 125M/170M [00:05<00:01, 32.8MB/s]
75%|#######5 | 128M/170M [00:05<00:01, 31.0MB/s]
77%|#######7 | 131M/170M [00:05<00:01, 29.3MB/s]
79%|#######9 | 135M/170M [00:06<00:01, 31.0MB/s]
81%|########1 | 138M/170M [00:06<00:01, 26.6MB/s]
83%|########2 | 141M/170M [00:06<00:01, 23.9MB/s]
84%|########4 | 143M/170M [00:06<00:01, 22.6MB/s]
85%|########5 | 145M/170M [00:06<00:01, 1
8.6MB/s]
87%|########6 | 147M/170M [00:06<00:01, 18.4MB/s]
89%|########8 | 151M/170M [00:06<00:00, 24.0MB/s]
91%|######### | 154M/170M [00:06<00:00, 26.8MB/s]
93%|#########2| 157M/170M [00:07<00:00, 25.6MB/s]
94%|#########4| 160M/170M [00:07<00:00, 25.5MB/s]
96%|#########6| 164M/170M [00:07<00:00, 29.5MB/s]
98%|#########8| 167M/170M [00:07<00:00, 29.1MB/s]
100%|#########9| 170M/170M [00:07<00:00, 25.6MB/s]
100%|##########| 170M/170M [00:07<00:00, 23.6MB/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').
@@ -253,7 +253,7 @@ Get boxes with score larger than 0.9
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 56.186 seconds)
+ **Total running time of the script:** ( 3 minutes 6.297 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 b94c5d96e..9c35a9169 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, 202MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 151MB/s]
@@ -344,7 +344,7 @@ 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)
- 88.8998 88.8076 90.4354 88.5657 0.2926
+ 90.0831 90.0874 90.3359 89.8714 0.1063
@@ -384,7 +384,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 2.748 seconds)
+ **Total running time of the script:** ( 1 minutes 2.913 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 7527eb031..52bbaebe1 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
@@ -351,7 +351,7 @@ 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)
- 120.9836 121.0695 122.7016 119.0857 0.6783
+ 115.7767 115.3796 119.0440 114.8219 1.1002
@@ -385,7 +385,7 @@ 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 55.046 seconds)
+ **Total running time of the script:** ( 1 minutes 58.369 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 0b7e8f548..8a6c3f2cf 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -221,7 +221,7 @@ We create a Relay VM to build and execute the model.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 8.987 seconds)
+ **Total running time of the script:** ( 1 minutes 13.060 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 6453c7b52..b31c6d115 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]
6%|5 | 7349/132723 [00:00<00:01, 73486.22KB/s]
12%|#2 | 16111/132723 [00:00<00:01, 81798.13KB/s]
19%|#8 | 24875/132723 [00:00<00:01, 84463.60KB/s]
25%|##5 | 33643/132723 [00:00<00:01, 85731.46KB/s]
32%|###1 | 42416/132723 [00:00<00:01, 86449.33KB/s]
39%|###8 | 51189/132723 [00:00<00:00, 86883.29KB/s]
45%|####5 | 60013/132723 [00:00<00:00, 87323.91KB/s]
52%|#####1 | 68746/132723 [00:00<00:00, 87270.04KB/s]
58%|#####8 | 77570/132723 [00:00<00:00, 87569.69KB/s]
65%|######5 | 86378/132723 [00:01<00:00, 87725.20KB/s]
72%|#######1 | 95244/132723 [00:01<00:00, 88010.03KB/s]
78%|#######8 | 104046/132723 [00:01<00:00, 87900.90KB/s]
85%|########5 | 112992/132723 [00:01<00:00, 88370.39KB/s]
92%|#########1| 121830/132723 [00:01<00:00, 88198.59KB/s]
98%|#########8| 130650/132723 [00:01<00:00, 88136.64KB/s]
100%|#######
###| 132723/132723 [00:01<00:00, 86921.94KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
4%|4 | 5745/132723 [00:00<00:02, 57445.54KB/s]
11%|#1 | 14701/132723 [00:00<00:01, 76333.91KB/s]
17%|#6 | 22335/132723 [00:00<00:01, 68895.87KB/s]
24%|##3 | 31418/132723 [00:00<00:01, 75736.95KB/s]
30%|### | 39974/132723 [00:00<00:01, 79098.77KB/s]
37%|###6 | 49075/132723 [00:00<00:01, 83009.14KB/s]
44%|####3 | 58055/132723 [00:00<00:00, 85174.55KB/s]
51%|##### | 67037/132723 [00:00<00:00, 86627.75KB/s]
57%|#####7 | 76096/132723 [00:00<00:00, 87850.70KB/s]
64%|######4 | 85214/132723 [00:01<00:00, 88869.63KB/s]
71%|#######1 | 94299/132723 [00:01<00:00, 89471.54KB/s]
78%|#######7 | 103259/132723 [00:01<00:00, 74528.01KB/s]
85%|########4 | 112382/132723 [00:01<00:00, 78952.81KB/s]
91%|######### | 120634/132723 [00:01<00:00, 72931.99KB/s]
98%|#########7| 129776/132723 [00:01<00:00, 77803.17KB/s]
100%|#######
###| 132723/132723 [00:01<00:00, 79789.63KB/s]
@@ -202,7 +202,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 23.182 seconds)
+ **Total running time of the script:** ( 2 minutes 17.880 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 36e666d75..033097c24 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
=================
-**10:14.287** total execution time for **how_to_deploy_models** files:
+**10:30.530** total execution time for **how_to_deploy_models** files:
-- **02:56.186**: :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``)
-- **02:23.182**: :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)
-- **01:55.046**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)
-- **01:08.987**: :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)
-- **01:02.748**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)
-- **00:27.220**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)
-- **00:20.718**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)
-- **00:00.200**: :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``)
+- **03:06.297**: :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``)
+- **02:17.880**: :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)
+- **01:58.369**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)
+- **01:13.060**: :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)
+- **01:02.913**: :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)
+- **00:29.066**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)
+- **00:22.768**: :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)
+- **00:00.178**: :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 ac07c0aff..e4cd911e3 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
@@ -423,7 +423,7 @@ 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.zip8180b182-e7c4-4fb2-b502-92a099f258e0 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip1ca18e90-79b0-4dd5-a7c0-cec1a87d0153 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 7d07da39e..5d1988ad3 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
=================
-**00:37.264** total execution time for **how_to_extend_tvm** files:
+**00:36.807** total execution time for **how_to_extend_tvm** files:
-- **00:33.819**: :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``)
-- **00:02.222**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)
-- **00:01.024**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)
-- **00:00.199**: :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``)
+- **00:33.410**: :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``)
+- **00:02.202**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)
+- **00:01.014**: :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)
+- **00:00.181**: :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 9e2bdb853..0c5ae1ae0 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: 5871us [5871us] (44.64%; 44.64%)
- FoldScaleAxis: 7280us [2us] (55.36%; 55.36%)
- FoldConstant: 7278us [1486us] (55.34%; 99.97%)
- InferType: 5792us [5792us] (44.04%; 79.58%)
+ InferType: 6129us [6129us] (45.43%; 45.43%)
+ FoldScaleAxis: 7361us [2us] (54.57%; 54.57%)
+ FoldConstant: 7359us [1507us] (54.55%; 99.97%)
+ InferType: 5851us [5851us] (43.38%; 79.51%)
@@ -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: 5805us [5805us] (45.05%; 45.05%)
- FoldScaleAxis: 7081us [2us] (54.95%; 54.95%)
- FoldConstant: 7080us [1457us] (54.94%; 99.97%)
- InferType: 5622us [5622us] (43.63%; 79.42%)
+ InferType: 5843us [5843us] (44.45%; 44.45%)
+ FoldScaleAxis: 7302us [2us] (55.55%; 55.55%)
+ FoldConstant: 7300us [1547us] (55.53%; 99.97%)
+ InferType: 5753us [5753us] (43.76%; 78.80%)
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 d4ea5d31e..1d40d0372 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: 54.231466 ms
+ Convolution: 54.154264 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 027c2ea4f..42a6ba1c3 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: 7.600894 ms
+ conv2d with tensor core: 6.776993 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 0419cf0e0..e3da1e506 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: 0.019028
- Baseline: 3.448080
+ Numpy running time: 0.018245
+ Baseline: 3.353084
@@ -210,7 +210,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.296260
+ Opt1: 0.292065
@@ -309,7 +309,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.338481
+ Opt2: 0.332295
@@ -401,7 +401,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.119919
+ Opt3: 0.112999
@@ -520,7 +520,7 @@ flattening.
.. code-block:: none
- Opt4: 0.111835
+ Opt4: 0.109010
@@ -638,7 +638,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.111540
+ Opt5: 0.110016
@@ -759,7 +759,7 @@ Futhermore, we can also utilize multi-core processors to do the thread-level par
.. code-block:: none
- Opt6: 0.145282
+ Opt6: 0.144609
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 898b9c3e5..ef947d0ad 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
=================
-**00:35.279** total execution time for **how_to_optimize_operators** files:
+**00:34.378** total execution time for **how_to_optimize_operators** files:
-- **00:32.586**: :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)
-- **00:01.446**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``)
-- **00:01.247**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)
+- **00:31.904**: :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)
+- **00:01.340**: :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``)
+- **00:01.134**: :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 4ead674e9..5b135307c 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
=================
-**05:03.892** total execution time for **how_to_tune_with_autoscheduler** files:
-
-- **02:22.015**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``)
-- **01:18.048**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)
-- **00:39.773**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)
-- **00:27.280**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)
-- **00:08.620**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)
-- **00:08.155**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)
+**03:34.635** total execution time for **how_to_tune_with_autoscheduler** files:
+
+- **01:20.332**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)
+- **01:08.042**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``)
+- **00:40.460**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)
+- **00:09.337**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)
+- **00:08.381**: :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)
+- **00:08.084**: :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 868a7991d..5dd17aafd 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
@@ -191,7 +191,6 @@ file and apply it.
-
We can lower the schedule to see the IR after auto-scheduling.
The auto-scheduler correctly performs optimizations including multi-level tiling,
cooperative fetching, unrolling and operator fusion.
@@ -222,491 +221,154 @@ 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" = 56;
- allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [168]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [1536]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope="local", align=8)[0] = 0f32
- conv2d_nchw_1[2] = 0f32
- conv2d_nchw_1[4] = 0f32
- conv2d_nchw_1[6] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 32;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [16]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [504]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [384]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [64], [], scope="local", align=32)[0] = 0f32
conv2d_nchw_1[8] = 0f32
- conv2d_nchw_1[10] = 0f32
- conv2d_nchw_1[12] = 0f32
conv2d_nchw_1[1] = 0f32
- conv2d_nchw_1[3] = 0f32
- conv2d_nchw_1[5] = 0f32
- conv2d_nchw_1[7] = 0f32
conv2d_nchw_1[9] = 0f32
+ conv2d_nchw_1[2] = 0f32
+ conv2d_nchw_1[10] = 0f32
+ conv2d_nchw_1[3] = 0f32
conv2d_nchw_1[11] = 0f32
+ conv2d_nchw_1[4] = 0f32
+ conv2d_nchw_1[12] = 0f32
+ conv2d_nchw_1[5] = 0f32
conv2d_nchw_1[13] = 0f32
+ conv2d_nchw_1[6] = 0f32
+ conv2d_nchw_1[14] = 0f32
+ conv2d_nchw_1[7] = 0f32
+ conv2d_nchw_1[15] = 0f32
for (rc.outer.outer: int32, 0, 64) {
for (ry.outer.outer: int32, 0, 3) {
- let cse_var_4: int32 = (rc.outer.outer*392)
- let cse_var_3: int32 = (ry.outer.outer*7)
- let cse_var_2: int32 = (rc.outer.outer*72)
+ let cse_var_3: int32 = (rc.outer.outer*392)
+ let cse_var_2: int32 = (ry.outer.outer*7)
let cse_var_1: int32 = (ry.outer.outer*3)
{
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [168], [], scope="shared")[(threadIdx.x_1*2)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1*2), 21), 3) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1*2), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)))) && ((floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)) < 8)), data[(((((cse_var_4 + (floordiv((threadIdx.x_1*2), 3)*7)) + cse_var_3) [...]
- pad_temp.shared_1[((threadIdx.x_1*2) + 1)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 1), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 1), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 1), 3)*7)) + cse_var_3) + floormod(blockIdx.x, 7 [...]
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [504], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 9) + ry.outer.outer)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[((((cse_var_3 + (floordiv(threadIdx.x_1, 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 49), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 49), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 49), 9)*7)) + cse_var_2) + 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 <= (floordiv(floormod((threadIdx.x_1 + 98), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 98), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 8), 9))) && (floormod((threadIdx.x_1 + 8), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 98), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 147), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 147), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 147), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 196), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 196), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 7), 9))) && (floormod((threadIdx.x_1 + 7), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 196), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 245), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 245), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 245), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 294), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 294), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 6), 9))) && (floormod((threadIdx.x_1 + 6), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 294), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 343), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 343), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 1), 9))) && (floormod((threadIdx.x_1 + 1), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 343), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 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 + 392)] = @tir.if_then_else(((((floordiv(floormod((threadIdx.x_1 + 392), 63), 9) + ry.outer.outer) < 8) && (1 <= floormod((threadIdx.x_1 + 5), 9))) && (floormod((threadIdx.x_1 + 5), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 392), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 5), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 9) + ry.outer.outer)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[((((cse_var_3 + (floordiv(threadIdx.x_1, 9)*7)) + cse_var_2) + floormod(threadIdx.x_1, 9)) + 335)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ if @tir.likely((threadIdx.x_1 < 14), dtype=bool) {
+ pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else(((((floordiv(floormod((threadIdx.x_1 + 490), 63), 9) + ry.outer.outer) < 8) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 490), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
}
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- pad_temp.shared_1[((threadIdx.x_1*2) + 64)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 64), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 64), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 64), 3)*7)) + cse_var_3) + floormod(blockIdx. [...]
- pad_temp.shared_1[((threadIdx.x_1*2) + 65)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 65), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 65), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 65), 3)*7)) + cse_var_3) + floormod(blockIdx. [...]
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ kernel.shared_1: Buffer(kernel.shared, float32, [384], [], scope="shared")[(threadIdx.x_2*4)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod((threadIdx.x_2*4), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 1), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 2), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 2), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv((threadIdx.x_2*4), 3) + 1), 8)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
}
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- if @tir.likely((threadIdx.x_1 < 20), dtype=bool) {
- pad_temp.shared_1[((threadIdx.x_1*2) + 128)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 128), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 128), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 128), 3)*7)) + cse_var_3) + floormod(blo [...]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 196), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
+ }
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 197), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 2), 3))]
+ }
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv((threadIdx.x_2*4), 3) + 2), 8)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
}
- if @tir.likely((threadIdx.x_1 < 20), dtype=bool) {
- pad_temp.shared_1[((threadIdx.x_1*2) + 129)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod((floordiv((threadIdx.x_1*2), 3) + 1), 7))) && ((ry.outer.outer + floormod((floordiv((threadIdx.x_1*2), 3) + 1), 7)) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)))) && ((floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)) < 8)), data[(((((cse_var_4 + (floordiv((threadIdx.x_1*2), 3)*7)) + cse_var_3) + floormod(blockIdx.x, 7)) + floormod((t [...]
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv(((threadIdx.x_2*4) + 196), 3) + 1), 8)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
+ }
+ }
+ for (rc.outer.inner: int32, 0, 2) {
+ for (rx.outer.inner: int32, 0, 3) {
+ let cse_var_4: int32 = ((rc.outer.inner*12) + rx.outer.inner)
+ {
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[cse_var_4]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 192)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 24)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 216)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 3)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 195)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 27)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 219)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 6)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 198)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 30)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 222)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 9)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 201)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 33)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 225)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 48)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 240)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 72)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 264)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 51)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 243)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 75)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 267)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 54)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 246)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 78)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 270)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 57)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 249)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 81)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 273)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 96)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 288)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 120)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 312)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 99)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 291)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 123)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 315)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 102)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 294)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 126)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 318)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 105)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 297)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 129)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 321)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 144)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 336)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 168)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 360)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 147)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 339)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 171)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 363)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 150)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 342)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 174)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 366)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 153)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 345)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 177)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 369)]))
+ }
}
}
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1: Buffer(kernel.shared, float32, [1536], [], scope="shared")[threadIdx.x_2] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 32)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 4), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 64)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 8), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 96)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 18432)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 16), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 160)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 20), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 192)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 224)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 28), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 32), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 288)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 55296)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 320)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 40), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 352)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 44), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 416)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 52), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 448)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 56), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 480)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 92160)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 64), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 544)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 68), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 576)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 608)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 76), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 80), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 672)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 129024)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 704)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 88), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 736)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 92), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 800)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 100), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 800), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 832)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 104), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 832), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 864)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 165888)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 112), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 896), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 928)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 116), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 928), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 960)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 992)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 124), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 992), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 128), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1024), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1056)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 202752)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 136), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1088), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1120)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 140), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1120), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1184)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 148), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1184), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 152), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1216), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1248)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 239616)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 160), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1280), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1312)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 164), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1312), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1376)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 172), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1376), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 176), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1408), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1440)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 276480)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 184), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1472), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1504)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 188), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1504), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[9]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[15]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[18]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*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) + 24)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*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) + 25)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[8]*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) + 26)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*48) + 47)]))
}
}
}
- for (i1.inner: int32, 0, 2) {
- compute[((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7))] = max((conv2d_nchw_1[i1.inner] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 7)] = max((conv2d_nchw_1[(i1.inner + 2)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 14)] = max((conv2d_nchw_1[(i1.inner + 4)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 21)] = max((conv2d_nchw_1[(i1.inner + 6)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 28)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 35)] = max((conv2d_nchw_1[(i1.inner + 10)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 42)] = max((conv2d_nchw_1[(i1.inner + 12)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
+ for (i1.inner: int32, 0, 8) {
+ compute[(((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias[((blockIdx.x*16) + i1.inner)]), 0f32)
+ compute[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 392)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias[(((blockIdx.x*16) + i1.inner) + 8)]), 0f32)
}
}
}
@@ -759,7 +421,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.348 ms
+ Execution time of this operator: 0.292 ms
@@ -803,36 +465,36 @@ 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=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=32)
- 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_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_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
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=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=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=7)
+ 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=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_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=1)
- conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=8)
+ 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=2)
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=3)
- conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_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)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2 [...]
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=2)
- compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=32)
- compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_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_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
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=1)
- compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=7)
+ compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+ 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=1)
+ compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
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)
@@ -850,16 +512,16 @@ 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=1)
+ 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=4)
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=32)
+ 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)
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=2)
+ 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)
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=32)
+ 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)
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", 512)
+ 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, "unroll_explicit", True)
CUDA source code:
@@ -877,432 +539,132 @@ 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__(32) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[14];
- __shared__ float pad_temp_shared[168];
- __shared__ float kernel_shared[1536];
+ 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[16];
+ __shared__ float pad_temp_shared[504];
+ __shared__ float kernel_shared[384];
conv2d_nchw[0] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
- conv2d_nchw[7] = 0.000000e+00f;
conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[13] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
+ conv2d_nchw[14] = 0.000000e+00f;
+ conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[15] = 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();
- pad_temp_shared[(((int)threadIdx.x) * 2)] = (((((1 <= ((((((int)threadIdx.x) * 2) % 21) / 3) + ry_outer_outer)) && (((((((int)threadIdx.x) * 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)))) && (((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 2) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) + ((((int)threadIdx.x) * 2) % 3)) - 8) [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 1)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 1) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) + [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 64)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 64) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 65)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 65) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) [...]
- if (((int)threadIdx.x) < 20) {
- pad_temp_shared[((((int)threadIdx.x) * 2) + 128)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 128) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % [...]
+ pad_temp_shared[((int)threadIdx.x)] = ((((1 <= ((((int)threadIdx.x) / 9) + ry_outer_outer)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 7)) + (ry_outer_outer * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((1 <= ((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 49) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 98)] = (((((1 <= ((((((int)threadIdx.x) + 35) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 35) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 98) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((1 <= ((((((int)threadIdx.x) + 21) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 21) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 147) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((1 <= ((((((int)threadIdx.x) + 7) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 7) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 196) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 245)] = (((((1 <= ((((((int)threadIdx.x) + 56) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 56) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 245) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 294)] = (((((1 <= ((((((int)threadIdx.x) + 42) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 42) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 294) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 343)] = (((((1 <= ((((((int)threadIdx.x) + 28) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 28) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 343) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((((((((int)threadIdx.x) + 14) % 63) / 9) + ry_outer_outer) < 8) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 392) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 <= ((((int)threadIdx.x) / 9) + ry_outer_outer)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 7)) + (ry_outer_outer * 7)) + (((int)threadIdx.x) % 9)) + 335)] : 0.000000e+00f);
+ if (((int)threadIdx.x) < 14) {
+ pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer) < 8) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 490) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
}
- if (((int)threadIdx.x) < 20) {
- pad_temp_shared[((((int)threadIdx.x) * 2) + 129)] = (((((1 <= (ry_outer_outer + ((((((int)threadIdx.x) * 2) / 3) + 1) % 7))) && ((ry_outer_outer + ((((((int)threadIdx.x) * 2) / 3) + 1) % 7)) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)))) && (((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 2) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) + ((((int)threadId [...]
+ kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) % 6) * 4) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) % 6) * 4) + 1) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) % 6) * 4) + 2) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 2) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) / 3) + 1) & 7) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) + 4) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) + 5) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 2) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) / 3) + 2) & 7) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + ((((((((int)threadIdx.x) * 4) + 196) / 3) + 1) & 7) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
}
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 32)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 32) / 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) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 96)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 18432)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 160)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 160) / 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) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 224)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 224) / 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) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 288)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 55296)];
- kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 352)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 352) / 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) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 416)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 416) / 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) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 480)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 92160)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 544)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 544) / 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) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 608)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 608) / 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) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 672)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 129024)];
- kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 736)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 736) / 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) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 800)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 800) / 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) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 864)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 165888)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 928)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 928) / 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) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 992)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 992) / 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) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1056)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 202752)];
- kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1120)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1120) / 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) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 1184)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1184) / 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) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1248)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 239616)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1312)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1312) / 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) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 1376)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1376) / 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) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1440)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 276480)];
- kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1504)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1504) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
__syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[9] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[15] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[18] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
+ for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[((rc_outer_inner * 12) + rx_outer_inner)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 192)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 216)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 3)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 195)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 219)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 6)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 198)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 222)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 9)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 201)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 225)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 48)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 240)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 72)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 264)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 51)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 243)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 75)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 267)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 54)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 246)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 78)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 270)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 57)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 249)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 81)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 273)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 96)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 288)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 120)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 312)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 99)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 291)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 123)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 315)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 102)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 294)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 126)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 318)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 105)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 297)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 129)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 321)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 144)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 336)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 168)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 360)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 147)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 339)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 171)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 363)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 150)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 342)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 174)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 366)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 153)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 345)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 177)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 369)]));
+ }
+ }
}
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[(((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7))] = max((conv2d_nchw[i1_inner] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 7)] = max((conv2d_nchw[(i1_inner + 2)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 14)] = max((conv2d_nchw[(i1_inner + 4)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 21)] = max((conv2d_nchw[(i1_inner + 6)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 28)] = max((conv2d_nchw[(i1_inner + 8)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 35)] = max((conv2d_nchw[(i1_inner + 10)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 42)] = max((conv2d_nchw[(i1_inner + 12)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 8; ++i1_inner) {
+ compute[(((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 16) + i1_inner)]), 0.000000e+00f);
+ compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 8)]), 0.000000e+00f);
}
}
@@ -1361,7 +723,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 2 minutes 22.015 seconds)
+ **Total running time of the script:** ( 1 minutes 8.042 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 9a78d5182..945f32297 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
@@ -614,7 +614,7 @@ 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)
- 9.6688 9.6681 9.7013 9.6371 0.0262
+ 9.8376 9.8295 9.9006 9.7827 0.0485
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 2e3957302..0322a70ad 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
@@ -633,7 +633,7 @@ 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)
- 773.9947 774.0448 778.1659 769.7735 3.4264
+ 764.8067 762.8136 771.0680 760.5385 4.5238
@@ -658,7 +658,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 18.048 seconds)
+ **Total running time of the script:** ( 1 minutes 20.332 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 e1857a562..1e87940e5 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
@@ -330,7 +330,6 @@ file and apply it.
-
We can lower the schedule to see the IR after auto-scheduling.
The auto-scheduler correctly performs optimizations including multi-level tiling,
layout transformation, parallelization, vectorization, unrolling, and operator fusion.
@@ -362,74 +361,30 @@ 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], []), placeholder_9: placeholder_16: Buffer(placeholder_14, float32, [128, 512], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_6: placeholder_19: Buffer(placeholder_11, float32, [4916, 16, 1], [])} {
- for (i0.outer.i1.outer.fused: int32, 0, 256) "parallel" {
- allocate(compute_4: Pointer(global float32), float32, [256]), storage_scope = global {
- for (i.inner.init: int32, 0, 16) {
- let cse_var_1: int32 = (i.inner.init*16)
- {
- compute_5: Buffer(compute_4, float32, [256], [])[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
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_5: placeholder_16: Buffer(placeholder_10, float32, [128, 256], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_9: placeholder_19: Buffer(placeholder_14, float32, [128, 512], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
+ for (i0.outer.i1.outer.fused: int32, 0, 64) "parallel" {
+ allocate(compute_4: Pointer(global float32), float32, [1024]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 32) {
+ for (j.init: int32, 0, 16) {
+ compute_5: Buffer(compute_4, float32, [1024], [])[(((i.outer.inner*512) + (i.inner.init*16)) + j.init)] = 0f32
+ }
}
- }
- for (elem_idx: int32, 0, let cse_var_2: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
- for (i.inner: int32, 0, 16) {
- let cse_var_21: int32 = floormod(i0.outer.i1.outer.fused, 32)
- let cse_var_20: int32 = (i.inner*16)
- let cse_var_19: int32 = (elem_idx*16)
- let cse_var_18: int32 = (cse_var_20 + 10)
- 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 = (cse_var_20 + 1)
- let cse_var_3: int32 = ((floordiv(i0.outer.i1.outer.fused, 32)*4096) + (i.inner*256))
- {
- compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[((placeholder_3[cse_var_21]*16) + cse_var_19)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 1)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 2)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 3)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 4)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 5)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 6)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 7)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 8)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 9)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 10)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 11)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 12)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 13)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 14)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 15)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
+ for (elem_idx: int32, 0, let cse_var_1: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_3[(cse_var_1 + 1)] - placeholder_3[cse_var_1])) {
+ for (i.inner: int32, 0, 32) {
+ for (j: int32, 0, 16) {
+ let cse_var_3: int32 = floormod(i0.outer.i1.outer.fused, 32)
+ let cse_var_2: int32 = (((i.outer.inner*512) + (i.inner*16)) + j)
+ compute_5[cse_var_2] = (compute_5[cse_var_2] + (placeholder_1[(((placeholder_3[cse_var_3]*16) + (elem_idx*16)) + j)]*max(placeholder[((((floordiv(i0.outer.i1.outer.fused, 32)*16384) + (i.outer.inner*8192)) + (i.inner*256)) + placeholder_2[(placeholder_3[cse_var_3] + elem_idx)])], 0f32)))
+ }
}
}
}
- for (i0.inner: int32, 0, 16) {
- let cse_var_22: int32 = (((floordiv(i0.outer.i1.outer.fused, 32)*8192) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16))
- compute[ramp(cse_var_22, 1, 16)] = max((compute_5[ramp((i0.inner*16), 1, 16)] + placeholder_4[ramp(cse_var_22, 1, 16)]), broadcast(0f32, 16))
+ for (i0.inner: int32, 0, 64) {
+ for (i1.inner: int32, 0, 16) {
+ let cse_var_4: int32 = ((((floordiv(i0.outer.i1.outer.fused, 32)*32768) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16)) + i1.inner)
+ compute[cse_var_4] = max((compute_5[((i0.inner*16) + i1.inner)] + placeholder_4[cse_var_4]), 0f32)
+ }
}
}
}
@@ -483,7 +438,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 1.936 ms
+ Execution time of this operator: 1.639 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 04e225df1..63294d87b 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
=================
-**00:44.450** total execution time for **how_to_tune_with_autotvm** files:
+**00:46.013** total execution time for **how_to_tune_with_autotvm** files:
-- **00:43.547**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)
-- **00:00.236**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)
-- **00:00.223**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)
-- **00:00.223**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.py``)
-- **00:00.220**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``)
+- **00:45.154**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)
+- **00:00.233**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)
+- **00:00.221**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)
+- **00:00.203**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``)
+- **00:00.201**: :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.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 91605ab4f..598e1f557 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: 111.03/111.03 result: MeasureResult(costs=(0.0020850042083333333,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8102667331695557, timestamp=1652131803.5408685) [('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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 6 GFLOPS: 92.44/92.44 result: MeasureResult(costs=(0.00250425925,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9339966773986816, timestamp=1652162349.4977636) [('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: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 8 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 9 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 10 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 11 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 12 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 13 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 14 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 15 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 16 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 17 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 18 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+ No: 19 GFLOPS: 0.00/92.44 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: 0x00007f8cd9001fa2
+ 12: 0x00007f3309641fa2
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: 143.78/143.78 result: MeasureResult(costs=(0.0016100564603174603,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1321747303009033, timestamp=1652131829.0709817) [('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: 143.45/143.45 result: MeasureResult(costs=(0.0016138067200000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3942267894744873, timestamp=1652162368.5986714) [('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: 0.001961
+ Time cost of this operator: 0.001996
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 3e194011a..e177d0056 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
@@ -292,10 +292,10 @@ 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 314.6 98.727 (1, 2, 10, 10, 3) 2 1
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.134 0.984 (1, 6, 10, 10) 1 1
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.923 0.29 (1, 1, 10, 10, 3) 1 1
- Total_time - 318.657 - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.9 98.761 (1, 2, 10, 10, 3) 2 1
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.953 (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 - 314.801 - - - -
@@ -357,10 +357,10 @@ 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 81.9 96.822 (1, 6, 10, 10, 1) 2 1
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.736 2.053 (1, 6, 10, 10) 1 1
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.952 1.125 (1, 1, 10, 10, 3) 1 1
- Total_time - 84.588 - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 79.35 96.737 (1, 6, 10, 10, 1) 2 1
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.77 2.158 (1, 6, 10, 10) 1 1
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.906 1.105 (1, 1, 10, 10, 3) 1 1
+ Total_time - 82.026 - - - -
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
index 9cc66b85a..33afe58d2 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_tflite.rst.txt
@@ -373,6 +373,7 @@ Now, compile the model for the target:
- ./codegen/host/src
- ./codegen/host/src/default_lib0.c
- ./codegen/host/src/default_lib1.c
+ - ./codegen/host/src/default_lib2.c
- ./executor-config
- ./executor-config/graph
- ./executor-config/graph/graph.json
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 e8a820174..eeb4a4729 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
=================
-**00:43.296** total execution time for **how_to_work_with_microtvm** files:
+**00:48.573** total execution time for **how_to_work_with_microtvm** files:
-- **00:39.264**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)
-- **00:03.445**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)
-- **00:00.199**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.py``)
-- **00:00.196**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.py``)
-- **00:00.192**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_reference_vm.py` (``micro_reference_vm.py``)
+- **00:44.401**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)
+- **00:03.621**: :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)
+- **00:00.193**: :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``)
+- **00:00.176**: :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 828914735..ec408799e 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
=================
-**00:08.757** total execution time for **how_to_work_with_relay** files:
+**00:05.141** total execution time for **how_to_work_with_relay** files:
-- **00:06.738**: :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)
-- **00:01.799**: :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)
-- **00:00.220**: :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``)
+- **00:03.501**: :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)
+- **00:01.439**: :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)
+- **00:00.201**: :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 37475694c..bca501c25 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
=================
-**00:05.700** total execution time for **how_to_work_with_schedules** files:
+**00:05.327** total execution time for **how_to_work_with_schedules** files:
-- **00:02.093**: :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)
-- **00:01.123**: :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)
-- **00:00.733**: :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)
-- **00:00.718**: :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)
-- **00:00.315**: :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)
-- **00:00.247**: :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``)
-- **00:00.243**: :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)
-- **00:00.228**: :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)
+- **00:02.061**: :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)
+- **00:00.877**: :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)
+- **00:00.715**: :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)
+- **00:00.711**: :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)
+- **00:00.291**: :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)
+- **00:00.227**: :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``)
+- **00:00.224**: :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``)
+- **00:00.221**: :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.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 7c03b90fa..ea374247a 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/tmp1iib50ex/input0.cc'\nsource_filename = \"/tmp/tmp1iib50ex/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/tmp3c3njbhh/input0.cc'\nsource_filename = \"/tmp/tmp3c3njbhh/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 d77dcde41..0c6dd531b 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
=================
-**00:19.815** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:19.724** total execution time for **topic_vta_tutorials_autotvm** files:
-- **00:19.615**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``)
-- **00:00.199**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)
+- **00:19.524**: :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``)
+- **00:00.200**: :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 7f577a6b8..d0325f6c4 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -265,7 +265,7 @@ 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 20.78s!
+ resnet18_v1 inference graph built in 20.51s!
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 2be09241e..177cb1e69 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -301,7 +301,7 @@ The compilation steps are:
/workspace/python/tvm/relay/build_module.py:431: DeprecationWarning: Please use input parameter mod (tvm.IRModule) instead of deprecated parameter mod (tvm.relay.function.Function)
DeprecationWarning,
- yolov3-tiny inference graph built in 14.57s!
+ yolov3-tiny inference graph built in 14.78s!
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 a1954e7a1..cade7e8d6 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
=================
-**01:26.168** total execution time for **topic_vta_tutorials_frontend** files:
+**01:27.925** total execution time for **topic_vta_tutorials_frontend** files:
-- **00:45.709**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)
-- **00:40.459**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``)
+- **00:47.148**: :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)
+- **00:40.777**: :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 bf112e012..888679ca3 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
=================
-**00:03.496** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.569** total execution time for **topic_vta_tutorials_optimize** files:
-- **00:02.957**: :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)
-- **00:00.539**: :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``)
+- **00:02.999**: :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)
+- **00:00.570**: :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 815a3d2d0..b6f8e40c0 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
=================
-**00:00.984** total execution time for **topic_vta_tutorials** files:
+**00:00.972** total execution time for **topic_vta_tutorials** files:
-- **00:00.498**: :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``)
-- **00:00.486**: :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``)
+- **00:00.491**: :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``)
+- **00:00.481**: :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 65ec3d942..c2769db05 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -184,7 +184,6 @@ trials, we can load the best schedule from the log file and apply it.
.. code-block:: none
- *E
@@ -306,7 +305,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 94.948 ms
+ Execution time of this operator: 93.394 ms
@@ -402,7 +401,7 @@ resume the status and do more 5 trials.
Resume search:
/usr/local/lib/python3.7/dist-packages/xgboost/training.py:17: UserWarning: Old style callback is deprecated. See: https://xgboost.readthedocs.io/en/latest/python/callbacks.html
warnings.warn(f'Old style callback is deprecated. See: {link}', UserWarning)
-
+ *E
@@ -415,11 +414,6 @@ Expression (TE) language that demonstrates how TVM can optimize computational
operations.
-.. rst-class:: sphx-glr-timing
-
- **Total running time of the script:** ( 1 minutes 14.991 seconds)
-
-
.. _sphx_glr_download_tutorial_auto_scheduler_matmul_x86.py:
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index 3dd470ea7..a5c7befdf 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -268,7 +268,7 @@ standard deviation.
.. code-block:: none
- {'mean': 495.4421859199988, 'median': 495.8792097000014, 'std': 1.2959323818152269}
+ {'mean': 492.68831364111975, 'median': 492.6090608467348, 'std': 0.26641364605524614}
@@ -482,32 +482,30 @@ the tuning data to.
.. code-block:: none
-
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 1/25] Current/Best: 19.85/ 23.23 GFLOPS | Progress: (4/10) | 5.08 s
[Task 1/25] Current/Best: 15.89/ 23.23 GFLOPS | Progress: (8/10) | 7.41 s
[Task 1/25] Current/Best: 17.96/ 23.23 GFLOPS | Progress: (10/10) | 8.26 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 2/25] Current/Best: 21.60/ 21.60 GFLOPS | Progress: (4/10) | 2.32 s
[Task 2/25] Current/Best: 13.25/ 21.60 GFLOPS | Progress: (8/10) | 3.60 s
[Task 2/25] Current/Best: 17.54/ 21.60 GFLOPS | Progress: (10/10) | 4.08 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 3/25] Current/Best: 17.85/ 17.85 GFLOPS | Progress: (4/10) | 2.77 s
[Task 3/25] Current/Best: 16.41/ 21.01 GFLOPS | Progress: (8/10) | 4.17 s
[Task 3/25] Current/Best: 16.88/ 21.01 GFLOPS | Progress: (10/10) | 5.17 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 4/25] Current/Best: 17.47/ 19.23 GFLOPS | Progress: (4/10) | 2.38 s
[Task 4/25] Current/Best: 9.98/ 19.23 GFLOPS | Progress: (8/10) | 8.53 s
[Task 4/25] Current/Best: 10.26/ 19.23 GFLOPS | Progress: (10/10) | 9.82 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 5/25] Current/Best: 20.86/ 20.86 GFLOPS | Progress: (4/10) | 2.55 s
[Task 5/25] Current/Best: 17.38/ 20.86 GFLOPS | Progress: (8/10) | 4.59 s
[Task 5/25] Current/Best: 8.19/ 20.86 GFLOPS | Progress: (10/10) | 5.63 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 6/25] Current/Best: 17.48/ 17.48 GFLOPS | Progress: (4/10) | 4.40 s
[Task 6/25] Current/Best: 16.11/ 22.71 GFLOPS | Progress: (8/10) | 6.43 s
[Task 6/25] Current/Best: 8.92/ 22.71 GFLOPS | Progress: (10/10) | 7.86 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 7/25] Current/Best: 12.08/ 16.05 GFLOPS | Progress: (4/10) | 4.15 s
[Task 7/25] Current/Best: 19.03/ 19.03 GFLOPS | Progress: (8/10) | 6.32 s
[Task 7/25] Current/Best: 17.05/ 19.03 GFLOPS | Progress: (10/10) | 7.99 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 8/25] Current/Best: 9.33/ 23.37 GFLOPS | Progress: (4/10) | 8.33 s
[Task 8/25] Current/Best: 8.82/ 23.37 GFLOPS | Progress: (8/10) | 15.68 s
[Task 8/25] Current/Best: 12.78/ 23.37 GFLOPS | Progress: (10/10) | 17.01 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 9/25] Current/Best: 17.54/ 17.54 GFLOPS | Progress: (4/10) | 2.65 s
[Task 9/25] Current/Best: 12.95/ 23.43 GFLOPS | Progress: (8/10) | 4.39 s
[Task 9/25] Current/Best: 11.84/ 23.43 GFLOPS | Progress: (10/10) | 8.29 s Done.
-
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 10/25] Current/Best: 18.33/ 18.33 GFLOPS | Progress: (4/10) | 2.44 s
[Task 10/25] Current/Best: 17.64/ 22.95 GFLOPS | Progress: (8/10) | 3.88 s
[Task 10/25] Current/Best: 14.98/ 22.95 GFLOPS | Progress: (10/10) | 4.87 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 11/25] Current/Best: 19.46/ 19.46 GFLOPS | Progress: (4/10) | 3.24 s
[Task 11/25] Current/Best: 18.14/ 19.46 GFLOPS | Progress: (8/10) | 5.04 s
[Task 11/25] Current/Best: 21.88/ 21.88 GFLOPS | Progress: (10/10) | 6.18 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 12/25] Current/Best: 9.09/ 17.72 GFLOPS | Progress: (4/10) | 2.91 s
[Task 12/25] Current/Best: 13.32/ 22.65 GFLOPS | Progress: (8/10) | 5.95 s
[Task 12/25] Current/Best: 13.19/ 22.65 GFLOPS | Progress: (10/10) | 8.06 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 13/25] Current/Best: 7.04/ 19.93 GFLOPS | Progress: (4/10) | 4.36 s
[Task 13/25] Current/Best: 11.53/ 19.93 GFLOPS | Progress: (8/10) | 6.76 s
[Task 13/25] Current/Best: 8.06/ 19.93 GFLOPS | Progress: (10/10) | 9.17 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 14/25] Current/Best: 20.99/ 20.99 GFLOPS | Progress: (4/10) | 5.34 s
[Task 14/25] Current/Best: 6.26/ 20.99 GFLOPS | Progress: (8/10) | 7.85 s
[Task 14/25] Current/Best: 11.95/ 20.99 GFLOPS | Progress: (10/10) | 9.35 s Done.
-
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 15/25] Current/Best: 7.05/ 13.70 GFLOPS | Progress: (4/10) | 5.00 s
[Task 15/25] Current/Best: 9.32/ 19.03 GFLOPS | Progress: (8/10) | 9.84 s
[Task 15/25] Current/Best: 21.29/ 21.29 GFLOPS | Progress: (10/10) | 10.46 s Done.
-
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 16/25] Current/Best: 13.51/ 21.27 GFLOPS | Progress: (4/10) | 2.49 s
[Task 16/25] Current/Best: 17.95/ 21.58 GFLOPS | Progress: (8/10) | 3.98 s
[Task 16/25] Current/Best: 17.97/ 21.58 GFLOPS | Progress: (10/10) | 4.56 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 17/25] Current/Best: 16.97/ 16.97 GFLOPS | Progress: (4/10) | 3.65 s
[Task 17/25] Current/Best: 16.76/ 20.30 GFLOPS | Progress: (8/10) | 5.61 s
[Task 17/25] Current/Best: 23.71/ 23.71 GFLOPS | Progress: (10/10) | 6.28 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 18/25] Current/Best: 4.03/ 10.44 GFLOPS | Progress: (4/10) | 4.82 s
[Task 18/25] Current/Best: 12.40/ 18.57 GFLOPS | Progress: (8/10) | 7.24 s
[Task 18/25] Current/Best: 10.42/ 18.57 GFLOPS | Progress: (10/10) | 8.27 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 19/25] Current/Best: 7.02/ 17.18 GFLOPS | Progress: (4/10) | 4.24 s
[Task 19/25] Current/Best: 9.59/ 19.48 GFLOPS | Progress: (8/10) | 7.07 s
[Task 19/25] Current/Best: 5.31/ 19.48 GFLOPS | Progress: (10/10) | 8.69 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 20/25] Current/Best: 17.78/ 17.78 GFLOPS | Progress: (4/10) | 2.98 s
[Task 20/25] Current/Best: 18.54/ 21.14 GFLOPS | Progress: (8/10) | 5.08 s
[Task 20/25] Current/Best: 21.42/ 21.42 GFLOPS | Progress: (10/10) | 5.95 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 21/25] Current/Best: 10.70/ 10.70 GFLOPS | Progress: (4/10) | 2.25 s Done.
-
[Task 21/25] Current/Best: 1.63/ 19.88 GFLOPS | Progress: (8/10) | 4.43 s
[Task 21/25] Current/Best: 6.70/ 19.88 GFLOPS | Progress: (10/10) | 5.65 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 22/25] Current/Best: 11.95/ 13.81 GFLOPS | Progress: (4/10) | 5.01 s
[Task 22/25] Current/Best: 11.14/ 13.81 GFLOPS | Progress: (8/10) | 7.89 s
[Task 22/25] Current/Best: 10.34/ 18.36 GFLOPS | Progress: (10/10) | 9.00 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 23/25] Current/Best: 20.39/ 20.39 GFLOPS | Progress: (4/10) | 2.91 s
[Task 23/25] Current/Best: 8.22/ 20.39 GFLOPS | Progress: (8/10) | 8.52 s
[Task 23/25] Current/Best: 10.71/ 20.39 GFLOPS | Progress: (10/10) | 9.75 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 24/25] Current/Best: 2.93/ 7.61 GFLOPS | Progress: (4/10) | 6.09 s
[Task 24/25] Current/Best: 11.07/ 11.07 GFLOPS | Progress: (8/10) | 7.07 s
[Task 24/25] Current/Best: 6.25/ 11.07 GFLOPS | Progress: (10/10) | 7.71 s Done.
-
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s Done.
-
[Task 25/25] Current/Best: 1.54/ 7.42 GFLOPS | Progress: (4/10) | 2.28 s
[Task 25/25] Current/Best: 5.68/ 9.51 GFLOPS | Progress: (8/10) | 4.55 s
[Task 25/25] Current/Best: 6.64/ 9.51 GFLOPS | Progress: (10/10) | 6.13 s Done.
-
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 1/25] Current/Best: 15.97/ 23.72 GFLOPS | Progress: (10/10) | 8.99 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 2/25] Current/Best: 3.97/ 18.83 GFLOPS | Progress: (10/10) | 6.13 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 3/25] Current/Best: 9.39/ 23.70 GFLOPS | Progress: (10/10) | 8.08 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 4/25] Current/Best: 11.41/ 23.24 GFLOPS | Progress: (10/10) | 5.04 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 5/25] Current/Best: 6.30/ 19.93 GFLOPS | Progress: (10/10) | 4.46 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 6/25] Current/Best: 15.94/ 22.46 GFLOPS | Progress: (10/10) | 5.21 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 7/25] Current/Best: 6.34/ 23.28 GFLOPS | Progress: (10/10) | 5.72 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 8/25] Current/Best: 8.82/ 22.72 GFLOPS | Progress: (10/10) | 5.71 s Done.
+
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 9/25] Current/Best: 13.90/ 23.89 GFLOPS | Progress: (10/10) | 13.89 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s Done.
+
[Task 10/25] Current/Best: 12.51/ 23.06 GFLOPS | Progress: (10/10) | 4.12 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 11/25] Current/Best: 12.82/ 23.12 GFLOPS | Progress: (10/10) | 6.06 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 12/25] Current/Best: 14.23/ 15.53 GFLOPS | Progress: (10/10) | 6.70 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 13/25] Current/Best: 6.97/ 19.98 GFLOPS | Progress: (10/10) | 6.87 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 14/25] Current/Best: 16.45/ 21.04 GFLOPS | Progress: (10/10) | 7.55 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 15/25] Current/Best: 19.41/ 23.14 GFLOPS | Progress: (10/10) | 7.21 s Done.
+
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 16/25] Current/Best: 16.65/ 20.23 GFLOPS | Progress: (10/10) | 4.97 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 17/25] Current/Best: 10.01/ 19.50 GFLOPS | Progress: (10/10) | 6.21 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 18/25] Current/Best: 9.04/ 23.38 GFLOPS | Progress: (10/10) | 6.53 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 19/25] Current/Best: 12.92/ 16.22 GFLOPS | Progress: (10/10) | 9.64 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 20/25] Current/Best: 10.08/ 17.87 GFLOPS | Progress: (10/10) | 7.75 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s Done.
+ Done.
+
[Task 21/25] Current/Best: 20.41/ 20.41 GFLOPS | Progress: (10/10) | 5.65 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 22/25] Current/Best: 10.59/ 22.61 GFLOPS | Progress: (10/10) | 6.68 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 23/25] Current/Best: 10.05/ 23.71 GFLOPS | Progress: (10/10) | 6.51 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s
[Task 24/25] Current/Best: 3.77/ 10.89 GFLOPS | Progress: (10/10) | 342.90 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/10) | 0.00 s Done.
+
[Task 25/25] Current/Best: 1.56/ 9.20 GFLOPS | Progress: (10/10) | 14.04 s
The output from this tuning process will look something like this:
@@ -595,8 +593,8 @@ Verify that the optimized model runs and produces the same results:
.. code-block:: none
- class='n02123045 tabby, tabby cat' with probability=0.621102
- class='n02123159 tiger cat' with probability=0.356379
+ class='n02123045 tabby, tabby cat' with probability=0.621104
+ class='n02123159 tiger cat' with probability=0.356378
class='n02124075 Egyptian cat' with probability=0.019712
class='n02129604 tiger, Panthera tigris' with probability=0.001215
class='n04040759 radiator' with probability=0.000262
@@ -649,8 +647,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 417.3321666199945, 'median': 416.87104789998557, 'std': 1.398795410530091}
- unoptimized: {'mean': 495.4421859199988, 'median': 495.8792097000014, 'std': 1.2959323818152269}
+ optimized: {'mean': 415.5365440319292, 'median': 415.2244868921116, 'std': 0.891752790513836}
+ unoptimized: {'mean': 492.68831364111975, 'median': 492.6090608467348, 'std': 0.26641364605524614}
@@ -670,7 +668,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 6 minutes 18.456 seconds)
+ **Total running time of the script:** ( 11 minutes 42.183 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 e93f09491..8df9571b7 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
- 1.287e-07 secs/op
+ 1.324e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index b8e55da73..da382025f 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, 0x25bd82b0)), stage(b, placeholder(b, 0x123e5470)), 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, 0x5dc8160)), stage(b, placeholder(b, 0x21bf9f20)), 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 336c08082..f8f0150fa 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
=================
-**09:27.581** total execution time for **tutorial** files:
+**14:25.948** total execution time for **tutorial** files:
-- **06:18.456**: :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)
-- **01:14.991**: :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``)
-- **00:59.509**: :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)
-- **00:26.639**: :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)
-- **00:25.854**: :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)
-- **00:01.060**: :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)
-- **00:00.701**: :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)
-- **00:00.202**: :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``)
-- **00:00.045**: :ref:`sphx_glr_tutorial_install.py` (``install.py``)
-- **00:00.045**: :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)
-- **00:00.041**: :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``)
-- **00:00.039**: :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``)
+- **11:42.183**: :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)
+- **01:00.571**: :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)
+- **00:56.544**: :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``)
+- **00:26.862**: :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)
+- **00:18.144**: :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)
+- **00:00.740**: :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)
+- **00:00.563**: :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)
+- **00:00.228**: :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``)
+- **00:00.031**: :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)
+- **00:00.028**: :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``)
+- **00:00.028**: :ref:`sphx_glr_tutorial_install.py` (``install.py``)
+- **00:00.027**: :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 ca0a2c0bc..a398ddb10 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -243,7 +243,7 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
- Numpy running time: 0.000007
+ Numpy running time: 0.000008
naive: 0.000006
@@ -438,10 +438,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 7.025479999356321e-06 1.0
- naive 5.8883e-06 0.8381349033147186
- parallel 6.0456e-06 0.8605248325458049
- vector 2.45482e-05 3.4941669469202274
+ numpy 8.097528479993344e-06 1.0
+ naive 5.8101e-06 0.717515228795438
+ parallel 6.0369e-06 0.7455237749290339
+ vector 2.4511299999999997e-05 3.0270100389965093
@@ -830,7 +830,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.017866
+ Numpy running time: 0.017551
@@ -886,7 +886,7 @@ optimizations.
.. code-block:: none
- none: 3.281898
+ none: 3.411887
@@ -985,7 +985,7 @@ schedule.
.. code-block:: none
- blocking: 0.299350
+ blocking: 0.293801
@@ -1077,7 +1077,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.336937
+ vectorization: 0.330251
@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], []),
@@ -1149,7 +1149,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.115925
+ loop permutation: 0.112472
@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], []),
@@ -1246,7 +1246,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.109349
+ array packing: 0.108019
@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], []),
@@ -1337,7 +1337,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.110721
+ block caching: 0.109893
@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], []),
@@ -1421,7 +1421,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.144684
+ parallelization: 0.144011
@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], []),
@@ -1500,13 +1500,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.2818980388000005 1.0
- blocking 0.29935010430000003 0.09121249373409997
- vectorization 0.33693663819999997 0.10266517552239317
- loop permutation 0.11592492090000002 0.035322523591374895
- array packing 0.10934909339999999 0.03331885759619229
- block caching 0.11072135629999999 0.033736988471611495
- parallelization 0.1446837584 0.04408539104185651
+ none 3.4118870004999997 1.0
+ blocking 0.2938010098 0.08611100243265517
+ vectorization 0.33025100120000006 0.09679423766133022
+ loop permutation 0.11247226430000001 0.03296482687835723
+ array packing 0.1080193138 0.031659698514097966
+ block caching 0.10989320740000001 0.03220892350300451
+ parallelization 0.1440112554 0.04220868257914042
@@ -1541,6 +1541,11 @@ 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 0.571 seconds)
+
+
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
diff --git a/docs/commit_hash b/docs/commit_hash
index f5e1832f2..cf272c43d 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-fb6eb38f311e820ec374113496294268f4160cdb
+8d4f4dd73ff2d762ade38aa0f9e8e9fba5f933da
diff --git a/docs/how_to/compile_models/from_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index 199f22bf4..1172ddb6e 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.zipf92e990a-8356-4320-b33c-7d175eacbac8 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.zipdd6354a5-a1b5-4986-b1e1-1c339d1b095f 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 02d3850ca..5567e2503 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -406,55 +406,49 @@ 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<07:31, 96.2kB/s]
- 0%| | 40.0k/41.5M [00:00<05:49, 124kB/s]
- 0%| | 88.0k/41.5M [00:00<03:38, 199kB/s]
- 0%| | 160k/41.5M [00:00<02:29, 290kB/s]
- 1%| | 336k/41.5M [00:00<01:16, 566kB/s]
- 1%|1 | 496k/41.5M [00:01<01:01, 699kB/s]
- 2%|2 | 0.98M/41.5M [00:01<00:29, 1.46MB/s]
- 4%|4 | 1.70M/41.5M [00:01<00:17, 2.40MB/s]
- 7%|6 | 2.85M/41.5M [00:01<00:13, 2.94MB/s]
- 10%|# | 4.32M/41.5M [00:01<00:08, 4.47MB/s]
- 13%|#3 | 5.48M/41.5M [00:02<00:07, 5.14MB/s]
- 14%|#4 | 5.99M/41.5M [00:02<00:08, 4.59MB/s]
- 16%|#6 | 6.66M/41.5M [00:02<00:08, 4.49MB/s]
- 19%|#9 | 7.95M/41.5M [00:02<00:08, 4.24MB/s]
- 23%|##2 | 9.42M/41.5M [00:02<00:06, 5.37MB/s]
- 25%|##5 | 10.4M/41.5M [00:03<00:05, 5.57MB/s]
- 28%|##7 | 11.4M/41.5M [00:03<00:05, 5.73MB/s]
- 30%|##9 | 12.4M/41.5M [00:03<00:05, 5.88MB/s]
- 32%|###2 | 13.5M/41.5M [00:03<00:04, 6.01MB/s]
- 35%|###5 | 14.5M/41.5M [00:03<00:04, 6.13MB/s]
- 38%|###7 | 15.6M/41.5M [00:03<00:04, 6.24MB/s]
- 40%|#### | 16.7M/41.5M [00:04<00:04, 6.33MB/s]
- 43%|####2 | 17.7M/41.5M [00:04<00:03, 6.41MB/s]
- 45%|####5 | 18.8M/41.5M [00:04<00:03, 6.95MB/s]
- 48%|####7 | 19.9M/41.5M [00:04<00:03, 7.44MB/s]
- 50%|##### | 20.9M/41.5M [00:04<00:02, 7.55MB/s]
- 52%|#####2 | 21.6M/41.5M [00:04<00:02, 7.08MB/s]
- 54%|#####3 | 22.3M/41.5M [00:04<00:03, 6.14MB/s]
- 56%|#####6 | 23.3M/41.5M [00:05<00:02, 6.66MB/s]
- 59%|#####8 | 24.4M/41.5M [00:05<00:02, 7.31MB/s]
- 61%|######1 | 25.4M/41.5M [00:05<00:02, 7.55MB/s]
- 63%|######2 | 26.1M/41.5M [00:05<00:02, 7.06MB/s]
- 65%|######4 | 26.8M/41.5M [00:05<00:02, 6.11MB/s]
- 67%|######7 | 27.9M/41.5M [00:05<00:02, 6.78MB/s]
- 70%|######9 | 29.0M/41.5M [00:05<00:01, 7.48MB/s]
- 72%|#######2 | 30.0M/41.5M [00:06<00:01, 8.16MB/s]
- 74%|#######4 | 30.8M/41.5M [00:06<00:01, 7.21MB/s]
- 76%|#######5 | 31.5M/41.5M [00:06<00:01, 6.28MB/s]
- 78%|#######8 | 32.5M/41.5M [00:06<00:01, 6.80MB/s]
- 81%|########1 | 33.7M/41.5M [00:06<00:01, 7.54MB/s]
- 84%|########3 | 34.7M/41.5M [00:06<00:00, 8.19MB/s]
- 86%|########5 | 35.5M/41.5M [00:06<00:00, 7.25MB/s]
- 87%|########7 | 36.2M/41.5M [00:07<00:00, 6.33MB/s]
- 90%|########9 | 37.3M/41.5M [00:07<00:00, 6.89MB/s]
- 93%|#########2| 38.4M/41.5M [00:07<00:00, 7.65MB/s]
- 95%|#########5| 39.5M/41.5M [00:07<00:00, 8.32MB/s]
- 97%|#########7| 40.3M/41.5M [00:07<00:00, 7.37MB/s]
- 99%|#########8| 41.0M/41.5M [00:07<00:00, 6.44MB/s]
-100%|##########| 41.5M/41.5M [00:07<00:00, 5.64MB/s]
+ 0%| | 16.0k/41.5M [00:00<07:42, 94.0kB/s]
+ 0%| | 48.0k/41.5M [00:00<04:52, 149kB/s]
+ 0%| | 96.0k/41.5M [00:00<03:27, 209kB/s]
+ 0%| | 168k/41.5M [00:00<02:28, 292kB/s]
+ 1%| | 344k/41.5M [00:00<01:17, 558kB/s]
+ 1%|1 | 560k/41.5M [00:01<00:53, 797kB/s]
+ 3%|2 | 1.09M/41.5M [00:01<00:26, 1.61MB/s]
+ 5%|4 | 1.99M/41.5M [00:01<00:14, 2.79MB/s]
+ 7%|7 | 3.00M/41.5M [00:01<00:10, 3.79MB/s]
+ 10%|# | 4.34M/41.5M [00:01<00:07, 5.13MB/s]
+ 14%|#4 | 5.81M/41.5M [00:01<00:05, 6.25MB/s]
+ 18%|#7 | 7.28M/41.5M [00:02<00:05, 7.02MB/s]
+ 21%|##1 | 8.74M/41.5M [00:02<00:03, 8.66MB/s]
+ 24%|##3 | 9.81M/41.5M [00:02<00:03, 9.23MB/s]
+ 26%|##5 | 10.8M/41.5M [00:02<00:03, 8.28MB/s]
+ 28%|##8 | 11.7M/41.5M [00:02<00:03, 7.92MB/s]
+ 32%|###1 | 13.1M/41.5M [00:02<00:03, 9.36MB/s]
+ 34%|###3 | 14.1M/41.5M [00:02<00:03, 9.08MB/s]
+ 36%|###6 | 15.0M/41.5M [00:02<00:03, 8.02MB/s]
+ 39%|###8 | 16.1M/41.5M [00:03<00:03, 8.10MB/s]
+ 42%|####2 | 17.5M/41.5M [00:03<00:02, 9.14MB/s]
+ 44%|####4 | 18.5M/41.5M [00:03<00:02, 9.26MB/s]
+ 47%|####6 | 19.4M/41.5M [00:03<00:02, 8.14MB/s]
+ 49%|####9 | 20.5M/41.5M [00:03<00:02, 7.71MB/s]
+ 53%|#####2 | 21.9M/41.5M [00:03<00:02, 8.61MB/s]
+ 56%|#####6 | 23.4M/41.5M [00:03<00:02, 9.44MB/s]
+ 59%|#####8 | 24.3M/41.5M [00:04<00:01, 9.49MB/s]
+ 61%|###### | 25.2M/41.5M [00:04<00:02, 8.33MB/s]
+ 63%|######3 | 26.3M/41.5M [00:04<00:01, 8.36MB/s]
+ 67%|######6 | 27.8M/41.5M [00:04<00:01, 9.31MB/s]
+ 69%|######9 | 28.7M/41.5M [00:04<00:01, 9.35MB/s]
+ 71%|#######1 | 29.6M/41.5M [00:04<00:01, 8.18MB/s]
+ 74%|#######4 | 30.7M/41.5M [00:04<00:01, 8.33MB/s]
+ 78%|#######7 | 32.2M/41.5M [00:05<00:01, 8.53MB/s]
+ 81%|########1 | 33.7M/41.5M [00:05<00:00, 8.57MB/s]
+ 85%|########4 | 35.1M/41.5M [00:05<00:00, 9.86MB/s]
+ 87%|########7 | 36.1M/41.5M [00:05<00:00, 9.41MB/s]
+ 89%|########9 | 37.0M/41.5M [00:05<00:00, 8.34MB/s]
+ 92%|#########1| 38.1M/41.5M [00:05<00:00, 8.53MB/s]
+ 95%|#########5| 39.5M/41.5M [00:05<00:00, 9.54MB/s]
+ 98%|#########7| 40.5M/41.5M [00:05<00:00, 9.11MB/s]
+100%|#########9| 41.3M/41.5M [00:06<00:00, 8.01MB/s]
+100%|##########| 41.5M/41.5M [00:06<00:00, 7.08MB/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 6ca126ba9..e8179a0f0 100644
--- a/docs/how_to/compile_models/from_paddle.html
+++ b/docs/how_to/compile_models/from_paddle.html
@@ -464,7 +464,7 @@ 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 13.489 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 3.912 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 c69d1b83a..0eca82b62 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -387,9 +387,11 @@ 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]
- 40%|###9 | 17.7M/44.7M [00:00<00:00, 185MB/s]
- 94%|#########3| 41.8M/44.7M [00:00<00:00, 225MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 222MB/s]
+ 5%|5 | 2.35M/44.7M [00:00<00:01, 24.3MB/s]
+ 12%|#1 | 5.17M/44.7M [00:00<00:01, 27.4MB/s]
+ 48%|####7 | 21.4M/44.7M [00:00<00:00, 92.2MB/s]
+ 82%|########2 | 36.7M/44.7M [00:00<00:00, 118MB/s]
+100%|##########| 44.7M/44.7M [00:00<00:00, 99.4MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_tensorflow.html b/docs/how_to/compile_models/from_tensorflow.html
index 5dd595095..ade534816 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -607,6 +607,7 @@ banana (score = 0.00022)
desk (score = 0.00019)
</pre></div>
</div>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 1.307 seconds)</p>
<div class="sphx-glr-footer class sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-tensorflow-py">
<div class="sphx-glr-download docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/7f1d3d1b878694c201c614c807cdebc8/from_tensorflow.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_tensorflow.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/sg_execution_times.html b/docs/how_to/compile_models/sg_execution_times.html
index 123aa2801..77befb65b 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>05:16.343</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>05:10.309</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<ul class="simple">
-<li><p><strong>01:13.489</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>00:58.690</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>00:54.078</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>00:30.502</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>00:25.434</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>00:21.068</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>00:20.620</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>00:17.985</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>00:12.122</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>00:02.355</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>01:03.912</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>01:01.307</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>00:55.356</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>00:29.339</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>00:24.634</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>00:21.462</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>00:20.545</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>00:19.195</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>00:11.792</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>00:02.766</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 2aa4fabcd..07bc4cc34 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -622,7 +622,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 15.3286 15.3173 15.4534 15.2190 0.0801
+ 15.5724 15.5598 15.6589 15.5272 0.0414
</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 7fcaf6fe4..a203ec739 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -409,14 +409,68 @@ 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]
- 10%|9 | 16.5M/170M [00:00<00:00, 174MB/s]
- 22%|##1 | 37.3M/170M [00:00<00:00, 199MB/s]
- 36%|###5 | 60.7M/170M [00:00<00:00, 220MB/s]
- 50%|####9 | 84.2M/170M [00:00<00:00, 231MB/s]
- 64%|######3 | 108M/170M [00:00<00:00, 238MB/s]
- 77%|#######7 | 131M/170M [00:00<00:00, 225MB/s]
- 90%|########9 | 153M/170M [00:00<00:00, 226MB/s]
-100%|##########| 170M/170M [00:00<00:00, 225MB/s]
+ 1%|1 | 1.88M/170M [00:00<00:11, 15.7MB/s]
+ 3%|3 | 5.31M/170M [00:00<00:07, 24.2MB/s]
+ 5%|4 | 8.24M/170M [00:00<00:06, 26.9MB/s]
+ 8%|7 | 13.0M/170M [00:00<00:04, 35.4MB/s]
+ 10%|# | 17.4M/170M [00:00<00:04, 39.0MB/s]
+ 12%|#2 | 21.2M/170M [00:00<00:04, 38.2MB/s]
+ 15%|#4 | 24.9M/170M [00:00<00:04, 37.4MB/s]
+ 17%|#6 | 28.5M/170M [00:00<00:05, 28.1MB/s]
+ 19%|#8 | 31.5M/170M [00:01<00:05, 25.7MB/s]
+ 20%|## | 34.1M/170M [00:01<00:07, 18.0MB/s]
+ 21%|##1 | 36.3M/170M [00:01<00:07, 18.3MB/s]
+ 23%|##2 | 38.5M/170M [00:01<00:07, 19.3MB/s]
+ 24%|##3 | 40.6M/170M [00:01<00:07, 18.6MB/s]
+ 25%|##5 | 42.6M/170M [00:01<00:07, 19.0MB/s]
+ 27%|##6 | 45.8M/170M [00:01<00:05, 22.7MB/s]
+ 28%|##8 | 48.1M/170M [00:02<00:05, 21.9MB/s]
+ 30%|##9 | 50.3M/170M [00:02<00:05, 21.9MB/s]
+ 31%|###1 | 53.1M/170M [00:02<00:05, 23.8MB/s]
+ 33%|###2 | 55.5M/170M [00:02<00:05, 23.8MB/s]
+ 34%|###4 | 57.8M/170M [00:02<00:05, 20.9MB/s]
+ 35%|###5 | 60.2M/170M [00:02<00:05, 22.1MB/s]
+ 38%|###7 | 63.8M/170M [00:02<00:04, 23.2MB/s]
+ 40%|#### | 68.6M/170M [00:02<00:03, 28.4MB/s]
+ 43%|####2 | 72.3M/170M [00:03<00:03, 28.6MB/s]
+ 44%|####4 | 75.0M/170M [00:03<00:03, 25.3MB/s]
+ 46%|####5 | 77.7M/170M [00:03<00:03, 25.9MB/s]
+ 48%|####7 | 80.8M/170M [00:03<00:03, 27.6MB/s]
+ 49%|####9 | 83.5M/170M [00:03<00:03, 25.2MB/s]
+ 51%|##### | 85.9M/170M [00:03<00:04, 21.8MB/s]
+ 52%|#####1 | 88.1M/170M [00:03<00:04, 18.5MB/s]
+ 53%|#####3 | 90.0M/170M [00:04<00:04, 18.2MB/s]
+ 54%|#####4 | 91.8M/170M [00:04<00:04, 18.3MB/s]
+ 55%|#####5 | 93.8M/170M [00:04<00:04, 18.7MB/s]
+ 56%|#####6 | 95.6M/170M [00:04<00:04, 17.3MB/s]
+ 57%|#####7 | 97.4M/170M [00:04<00:04, 17.6MB/s]
+ 58%|#####8 | 99.1M/170M [00:04<00:04, 17.8MB/s]
+ 59%|#####9 | 101M/170M [00:04<00:04, 17.2MB/s]
+ 60%|###### | 103M/170M [00:04<00:04, 16.1MB/s]
+ 61%|######1 | 104M/170M [00:04<00:04, 14.5MB/s]
+ 62%|######2 | 106M/170M [00:05<00:04, 15.7MB/s]
+ 64%|######3 | 108M/170M [00:05<00:03, 16.9MB/s]
+ 65%|######4 | 110M/170M [00:05<00:03, 19.4MB/s]
+ 67%|######6 | 114M/170M [00:05<00:02, 23.1MB/s]
+ 69%|######9 | 118M/170M [00:05<00:01, 29.2MB/s]
+ 71%|#######1 | 121M/170M [00:05<00:01, 29.4MB/s]
+ 73%|#######3 | 125M/170M [00:05<00:01, 32.8MB/s]
+ 75%|#######5 | 128M/170M [00:05<00:01, 31.0MB/s]
+ 77%|#######7 | 131M/170M [00:05<00:01, 29.3MB/s]
+ 79%|#######9 | 135M/170M [00:06<00:01, 31.0MB/s]
+ 81%|########1 | 138M/170M [00:06<00:01, 26.6MB/s]
+ 83%|########2 | 141M/170M [00:06<00:01, 23.9MB/s]
+ 84%|########4 | 143M/170M [00:06<00:01, 22.6MB/s]
+ 85%|########5 | 145M/170M [00:06<00:01, 18.6MB/s]
+ 87%|########6 | 147M/170M [00:06<00:01, 18.4MB/s]
+ 89%|########8 | 151M/170M [00:06<00:00, 24.0MB/s]
+ 91%|######### | 154M/170M [00:06<00:00, 26.8MB/s]
+ 93%|#########2| 157M/170M [00:07<00:00, 25.6MB/s]
+ 94%|#########4| 160M/170M [00:07<00:00, 25.5MB/s]
+ 96%|#########6| 164M/170M [00:07<00:00, 29.5MB/s]
+ 98%|#########8| 167M/170M [00:07<00:00, 29.1MB/s]
+100%|#########9| 170M/170M [00:07<00:00, 25.6MB/s]
+100%|##########| 170M/170M [00:07<00:00, 23.6MB/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').
@@ -509,7 +563,7 @@ 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 56.186 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 6.297 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 0c2964c81..7e56611a8 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, 202MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 151MB/s]
</pre></div>
</div>
</div>
@@ -539,7 +539,7 @@ 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)
- 88.8998 88.8076 90.4354 88.5657 0.2926
+ 90.0831 90.0874 90.3359 89.8714 0.1063
</pre></div>
</div>
<div class="admonition note">
@@ -578,7 +578,7 @@ 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 2.748 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 2.913 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 3abc3ec24..0ca005d9d 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -540,7 +540,7 @@ 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)
- 120.9836 121.0695 122.7016 119.0857 0.6783
+ 115.7767 115.3796 119.0440 114.8219 1.1002
</pre></div>
</div>
<div class="admonition note">
@@ -568,7 +568,7 @@ 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 55.046 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 58.369 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 c792c0f98..e994c189d 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -480,7 +480,7 @@ 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 8.987 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 13.060 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 c91faf715..e3d6a20aa 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]
- 6%|5 | 7349/132723 [00:00<00:01, 73486.22KB/s]
- 12%|#2 | 16111/132723 [00:00<00:01, 81798.13KB/s]
- 19%|#8 | 24875/132723 [00:00<00:01, 84463.60KB/s]
- 25%|##5 | 33643/132723 [00:00<00:01, 85731.46KB/s]
- 32%|###1 | 42416/132723 [00:00<00:01, 86449.33KB/s]
- 39%|###8 | 51189/132723 [00:00<00:00, 86883.29KB/s]
- 45%|####5 | 60013/132723 [00:00<00:00, 87323.91KB/s]
- 52%|#####1 | 68746/132723 [00:00<00:00, 87270.04KB/s]
- 58%|#####8 | 77570/132723 [00:00<00:00, 87569.69KB/s]
- 65%|######5 | 86378/132723 [00:01<00:00, 87725.20KB/s]
- 72%|#######1 | 95244/132723 [00:01<00:00, 88010.03KB/s]
- 78%|#######8 | 104046/132723 [00:01<00:00, 87900.90KB/s]
- 85%|########5 | 112992/132723 [00:01<00:00, 88370.39KB/s]
- 92%|#########1| 121830/132723 [00:01<00:00, 88198.59KB/s]
- 98%|#########8| 130650/132723 [00:01<00:00, 88136.64KB/s]
-100%|##########| 132723/132723 [00:01<00:00, 86921.94KB/s]
+ 4%|4 | 5745/132723 [00:00<00:02, 57445.54KB/s]
+ 11%|#1 | 14701/132723 [00:00<00:01, 76333.91KB/s]
+ 17%|#6 | 22335/132723 [00:00<00:01, 68895.87KB/s]
+ 24%|##3 | 31418/132723 [00:00<00:01, 75736.95KB/s]
+ 30%|### | 39974/132723 [00:00<00:01, 79098.77KB/s]
+ 37%|###6 | 49075/132723 [00:00<00:01, 83009.14KB/s]
+ 44%|####3 | 58055/132723 [00:00<00:00, 85174.55KB/s]
+ 51%|##### | 67037/132723 [00:00<00:00, 86627.75KB/s]
+ 57%|#####7 | 76096/132723 [00:00<00:00, 87850.70KB/s]
+ 64%|######4 | 85214/132723 [00:01<00:00, 88869.63KB/s]
+ 71%|#######1 | 94299/132723 [00:01<00:00, 89471.54KB/s]
+ 78%|#######7 | 103259/132723 [00:01<00:00, 74528.01KB/s]
+ 85%|########4 | 112382/132723 [00:01<00:00, 78952.81KB/s]
+ 91%|######### | 120634/132723 [00:01<00:00, 72931.99KB/s]
+ 98%|#########7| 129776/132723 [00:01<00:00, 77803.17KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 79789.63KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -470,7 +470,7 @@ 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 23.182 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 17.880 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 f4bcaaf6b..23254053d 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>10:14.287</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>10:30.530</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<ul class="simple">
-<li><p><strong>02:56.186</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>02:23.182</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>01:55.046</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>01:08.987</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>01:02.748</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>00:27.220</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>00:20.718</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>00:00.200</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>03:06.297</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>02:17.880</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>01:58.369</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>01:13.060</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>01:02.913</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>00:29.066</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>00:22.768</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>00:00.178</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 3825f1a54..114d2e3e1 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -588,7 +588,7 @@ 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.zip8180b182-e7c4-4fb2-b502-92a099f258e0 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.zip1ca18e90-79b0-4dd5-a7c0-cec1a87d0153 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 8321f900b..af4b32dc1 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>00:37.264</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:36.807</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:33.819</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>00:02.222</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>00:01.024</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>00:00.199</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>00:33.410</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>00:02.202</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>00:01.014</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>00:00.181</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 b31e0f6e3..e2f17f9e6 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: 5871us [5871us] (44.64%; 44.64%)
-FoldScaleAxis: 7280us [2us] (55.36%; 55.36%)
- FoldConstant: 7278us [1486us] (55.34%; 99.97%)
- InferType: 5792us [5792us] (44.04%; 79.58%)
+InferType: 6129us [6129us] (45.43%; 45.43%)
+FoldScaleAxis: 7361us [2us] (54.57%; 54.57%)
+ FoldConstant: 7359us [1507us] (54.55%; 99.97%)
+ InferType: 5851us [5851us] (43.38%; 79.51%)
</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: 5805us [5805us] (45.05%; 45.05%)
-FoldScaleAxis: 7081us [2us] (54.95%; 54.95%)
- FoldConstant: 7080us [1457us] (54.94%; 99.97%)
- InferType: 5622us [5622us] (43.63%; 79.42%)
+InferType: 5843us [5843us] (44.45%; 44.45%)
+FoldScaleAxis: 7302us [2us] (55.55%; 55.55%)
+ FoldConstant: 7300us [1547us] (55.53%; 99.97%)
+ InferType: 5753us [5753us] (43.76%; 78.80%)
</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 8d83e1cd5..88717bf80 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: 54.231466 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.154264 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 978bc7426..ce1fceded 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: 7.600894 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 6.776993 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 93a7dd099..221cb2e05 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: 0.019028
-Baseline: 3.448080
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018245
+Baseline: 3.353084
</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: 0.296260
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.292065
</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: 0.338481
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.332295
</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: 0.119919
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.112999
</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: 0.111835
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109010
</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: 0.111540
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.110016
</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: 0.145282
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.144609
</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 877f7afca..beb6088b8 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>00:35.279</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:34.378</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:32.586</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>00:01.446</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>00:01.247</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>00:31.904</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>00:01.340</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>00:01.134</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 826b0d7d0..ac495fd9f 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>05:03.892</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>03:34.635</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<ul class="simple">
-<li><p><strong>02:22.015</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>01:18.048</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>00:39.773</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>00:27.280</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>00:08.620</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>00:08.155</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>01:20.332</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>01:08.042</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>00:40.460</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>00:09.337</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>00:08.381</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>00:08.084</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 826d3bf9e..137ce1142 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
@@ -470,491 +470,154 @@ 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" = 56;
- allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [168]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [1536]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [4], [], scope="local", align=8)[0] = 0f32
- conv2d_nchw_1[2] = 0f32
- conv2d_nchw_1[4] = 0f32
- conv2d_nchw_1[6] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 32;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [16]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [504]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [384]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [64], [], scope="local", align=32)[0] = 0f32
conv2d_nchw_1[8] = 0f32
- conv2d_nchw_1[10] = 0f32
- conv2d_nchw_1[12] = 0f32
conv2d_nchw_1[1] = 0f32
- conv2d_nchw_1[3] = 0f32
- conv2d_nchw_1[5] = 0f32
- conv2d_nchw_1[7] = 0f32
conv2d_nchw_1[9] = 0f32
+ conv2d_nchw_1[2] = 0f32
+ conv2d_nchw_1[10] = 0f32
+ conv2d_nchw_1[3] = 0f32
conv2d_nchw_1[11] = 0f32
+ conv2d_nchw_1[4] = 0f32
+ conv2d_nchw_1[12] = 0f32
+ conv2d_nchw_1[5] = 0f32
conv2d_nchw_1[13] = 0f32
+ conv2d_nchw_1[6] = 0f32
+ conv2d_nchw_1[14] = 0f32
+ conv2d_nchw_1[7] = 0f32
+ conv2d_nchw_1[15] = 0f32
for (rc.outer.outer: int32, 0, 64) {
for (ry.outer.outer: int32, 0, 3) {
- let cse_var_4: int32 = (rc.outer.outer*392)
- let cse_var_3: int32 = (ry.outer.outer*7)
- let cse_var_2: int32 = (rc.outer.outer*72)
+ let cse_var_3: int32 = (rc.outer.outer*392)
+ let cse_var_2: int32 = (ry.outer.outer*7)
let cse_var_1: int32 = (ry.outer.outer*3)
{
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [168], [], scope="shared")[(threadIdx.x_1*2)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1*2), 21), 3) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1*2), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)))) && ((floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)) < 8)), data[(((((cse_var_4 + (floor [...]
- pad_temp.shared_1[((threadIdx.x_1*2) + 1)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 1), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 1), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 1), 3)*7)) + cse [...]
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [504], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 9) + ry.outer.outer)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[((((cse_var_3 + (floordiv(threadIdx.x_1, 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 49), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 49), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 49), 9)*7)) + cse_var_2) + 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 <= (floordiv(floormod((threadIdx.x_1 + 98), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 98), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 8), 9))) && (floormod((threadIdx.x_1 + 8), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 98), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 147), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 147), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 147), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 196), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 196), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 7), 9))) && (floormod((threadIdx.x_1 + 7), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 196), 9)*7)) + cse_var_2) + 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(((((1 <= (floordiv(floormod((threadIdx.x_1 + 245), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 245), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 245), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 2), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 294)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 294), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 294), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 6), 9))) && (floormod((threadIdx.x_1 + 6), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 294), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 6), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 343)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 343), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1 + 343), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1 + 1), 9))) && (floormod((threadIdx.x_1 + 1), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 343), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 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 + 392)] = @tir.if_then_else(((((floordiv(floormod((threadIdx.x_1 + 392), 63), 9) + ry.outer.outer) < 8) && (1 <= floormod((threadIdx.x_1 + 5), 9))) && (floormod((threadIdx.x_1 + 5), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 392), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 5), 9)) - 8)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ pad_temp.shared_1[(threadIdx.x_1 + 441)] = @tir.if_then_else((((1 <= (floordiv(threadIdx.x_1, 9) + ry.outer.outer)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data[((((cse_var_3 + (floordiv(threadIdx.x_1, 9)*7)) + cse_var_2) + floormod(threadIdx.x_1, 9)) + 335)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49;
+ if @tir.likely((threadIdx.x_1 < 14), dtype=bool) {
+ pad_temp.shared_1[(threadIdx.x_1 + 490)] = @tir.if_then_else(((((floordiv(floormod((threadIdx.x_1 + 490), 63), 9) + ry.outer.outer) < 8) && (1 <= floormod((threadIdx.x_1 + 4), 9))) && (floormod((threadIdx.x_1 + 4), 9) < 8)), data[((((cse_var_3 + (floordiv((threadIdx.x_1 + 490), 9)*7)) + cse_var_2) + floormod((threadIdx.x_1 + 4), 9)) - 8)], 0f32, dtype=float32)
}
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- pad_temp.shared_1[((threadIdx.x_1*2) + 64)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 64), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 64), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 1), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 64), 3)*7)) + [...]
- pad_temp.shared_1[((threadIdx.x_1*2) + 65)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 65), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 65), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 65), 3)*7)) + [...]
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ kernel.shared_1: Buffer(kernel.shared, float32, [384], [], scope="shared")[(threadIdx.x_2*4)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod((threadIdx.x_2*4), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 1)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 1), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 2)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 2), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 2), 3))]
+ kernel.shared_1[((threadIdx.x_2*4) + 3)] = kernel[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv((threadIdx.x_2*4), 3) + 1), 8)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
}
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32 {
- if @tir.likely((threadIdx.x_1 < 20), dtype=bool) {
- pad_temp.shared_1[((threadIdx.x_1*2) + 128)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*2) + 128), 21), 3) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*2) + 128), 21), 3) + ry.outer.outer) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)))) && ((floormod(blockIdx.x, 7) + floormod(((threadIdx.x_1*2) + 2), 3)) < 8)), data[(((((cse_var_4 + (floordiv(((threadIdx.x_1*2) + 128), 3) [...]
+ attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 49 {
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 196)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 196), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
}
- if @tir.likely((threadIdx.x_1 < 20), dtype=bool) {
- pad_temp.shared_1[((threadIdx.x_1*2) + 129)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod((floordiv((threadIdx.x_1*2), 3) + 1), 7))) && ((ry.outer.outer + floormod((floordiv((threadIdx.x_1*2), 3) + 1), 7)) < 8)) && (1 <= (floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)))) && ((floormod(blockIdx.x, 7) + floormod((threadIdx.x_1*2), 3)) < 8)), data[(((((cse_var_4 + (floordiv((threadIdx.x_1*2), 3)*7)) + cse_var_3) + floorm [...]
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 197)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floordiv(floormod(((threadIdx.x_2*4) + 197), 24), 3)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 2), 3))]
+ }
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 198)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv((threadIdx.x_2*4), 3) + 2), 8)*9)) + cse_var_1) + floormod((threadIdx.x_2*4), 3))]
+ }
+ if @tir.likely((threadIdx.x_2 < 47), dtype=bool) {
+ kernel.shared_1[((threadIdx.x_2*4) + 199)] = kernel[((((((blockIdx.x*73728) + (floordiv((threadIdx.x_2 + 49), 6)*4608)) + (rc.outer.outer*72)) + (floormod((floordiv(((threadIdx.x_2*4) + 196), 3) + 1), 8)*9)) + cse_var_1) + floormod(((threadIdx.x_2*4) + 1), 3))]
+ }
+ }
+ for (rc.outer.inner: int32, 0, 2) {
+ for (rx.outer.inner: int32, 0, 3) {
+ let cse_var_4: int32 = ((rc.outer.inner*12) + rx.outer.inner)
+ {
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[cse_var_4]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 192)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 24)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 216)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 3)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 195)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 27)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 219)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 6)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 198)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 30)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 222)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 9)]))
+ conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 201)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 33)]))
+ conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 225)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 48)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 240)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 72)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 264)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 51)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 243)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 75)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 267)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 54)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 246)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 78)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 270)]))
+ conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 57)]))
+ conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 249)]))
+ conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 81)]))
+ conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 273)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 96)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 288)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 120)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 312)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 99)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 291)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 123)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 315)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 102)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 294)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 126)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 318)]))
+ conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 105)]))
+ conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 297)]))
+ conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 129)]))
+ conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 321)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 144)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 336)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 168)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7))]*kernel.shared_1[(cse_var_4 + 360)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 147)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 339)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 171)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(cse_var_4 + 363)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 150)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 342)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 174)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 126)]*kernel.shared_1[(cse_var_4 + 366)]))
+ conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 153)]))
+ conv2d_nchw_1[14] = (conv2d_nchw_1[14] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 345)]))
+ conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 177)]))
+ conv2d_nchw_1[15] = (conv2d_nchw_1[15] + (pad_temp.shared_1[(((((rc.outer.inner*252) + (floordiv(threadIdx.x, 7)*9)) + rx.outer.inner) + floormod(threadIdx.x, 7)) + 189)]*kernel.shared_1[(cse_var_4 + 369)]))
+ }
}
}
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1: Buffer(kernel.shared, float32, [1536], [], scope="shared")[threadIdx.x_2] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 32)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 4), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 64)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 8), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 96)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 18432)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 16), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 160)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 20), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 192)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 224)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 28), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 32), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 288)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 55296)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 320)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 40), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 352)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 44), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 416)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 52), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 448)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 56), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 480)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 92160)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 64), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 544)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 68), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 576)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 608)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 76), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 80), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 672)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 129024)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 704)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 88), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 736)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 92), 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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 800)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 100), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 800), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 832)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 104), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 832), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 864)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 165888)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 112), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 896), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 928)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 116), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 928), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 960)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 992)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 124), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 992), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 128), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1024), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1056)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 202752)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 136), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1088), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1120)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 140), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1120), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1184)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 148), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1184), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 152), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1216), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1248)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 239616)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 160), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1280), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1312)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 164), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1312), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1376)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 172), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1376), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 176), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1408), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1440)] = kernel[(((((((floordiv(blockIdx.x, 7)*294912) + (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)) + 276480)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 184), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1472), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 32;
- kernel.shared_1[(threadIdx.x_2 + 1504)] = kernel[((((((floordiv(blockIdx.x, 7)*294912) + (floordiv((floordiv(threadIdx.x_2, 8) + 188), 3)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 1504), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[9]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[15]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[18]*kernel.shared_1[(threadIdx.x*48)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*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) + 24)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*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) + 25)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 25)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[8]*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) + 26)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 26)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 27)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 28)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 29)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 30)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 31)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 32)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*48) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*48) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*48) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*48) + 33)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*48) + 34)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*48) + 35)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*48) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*48) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*48) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*48) + 36)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*48) + 37)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*48) + 38)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*48) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*48) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*48) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*48) + 39)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*48) + 40)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*48) + 41)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*48) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*48) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*48) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*48) + 42)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*48) + 43)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*48) + 44)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*48) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*48) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*48) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*48) + 45)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*48) + 46)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*48) + 47)]))
- conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*48) + 47)]))
}
}
}
- for (i1.inner: int32, 0, 2) {
- compute[((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7))] = max((conv2d_nchw_1[i1.inner] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 7)] = max((conv2d_nchw_1[(i1.inner + 2)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 14)] = max((conv2d_nchw_1[(i1.inner + 4)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 21)] = max((conv2d_nchw_1[(i1.inner + 6)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 28)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 35)] = max((conv2d_nchw_1[(i1.inner + 10)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
- compute[(((((floordiv(blockIdx.x, 7)*3136) + (threadIdx.x*98)) + (i1.inner*49)) + floormod(blockIdx.x, 7)) + 42)] = max((conv2d_nchw_1[(i1.inner + 12)] + bias[(((floordiv(blockIdx.x, 7)*64) + (threadIdx.x*2)) + i1.inner)]), 0f32)
+ for (i1.inner: int32, 0, 8) {
+ compute[(((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x)] = max((conv2d_nchw_1[i1.inner] + bias[((blockIdx.x*16) + i1.inner)]), 0f32)
+ compute[((((blockIdx.x*784) + (i1.inner*49)) + threadIdx.x) + 392)] = max((conv2d_nchw_1[(i1.inner + 8)] + bias[(((blockIdx.x*16) + i1.inner) + 8)]), 0f32)
}
}
}
@@ -992,7 +655,7 @@ 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: 0.348 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.292 ms
</pre></div>
</div>
</div>
@@ -1022,36 +685,36 @@ 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=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=32)
-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_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_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
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=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=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=7)
+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=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_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=1)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=8)
+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=2)
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=3)
-conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_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)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, 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=2)
-compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=32)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_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_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
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=1)
-compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=7)
+compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+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=1)
+compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
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)
@@ -1069,16 +732,16 @@ 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=1)
+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=4)
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=32)
+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)
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=2)
+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)
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=32)
+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)
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", 512)
+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, "unroll_explicit", True)
CUDA source code:
@@ -1096,432 +759,132 @@ CUDA source code:
#define int64_t long long
#define uint64_t unsigned long long
#endif
-extern "C" __global__ void __launch_bounds__(32) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[14];
- __shared__ float pad_temp_shared[168];
- __shared__ float kernel_shared[1536];
+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[16];
+ __shared__ float pad_temp_shared[504];
+ __shared__ float kernel_shared[384];
conv2d_nchw[0] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
- conv2d_nchw[7] = 0.000000e+00f;
conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[13] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
+ conv2d_nchw[14] = 0.000000e+00f;
+ conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[15] = 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();
- pad_temp_shared[(((int)threadIdx.x) * 2)] = (((((1 <= ((((((int)threadIdx.x) * 2) % 21) / 3) + ry_outer_outer)) && (((((((int)threadIdx.x) * 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)))) && (((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 2) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)blockIdx.x) % 7)) + (((( [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 1)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 1) / 3) * 7)) + (ry_outer_outer * 7 [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 64)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 1) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 1) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 64) / 3) * 7)) + (ry_outer_outer * [...]
- pad_temp_shared[((((int)threadIdx.x) * 2) + 65)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 65) / 3) * 7)) + (ry_outer_outer * [...]
- if (((int)threadIdx.x) < 20) {
- pad_temp_shared[((((int)threadIdx.x) * 2) + 128)] = (((((1 <= (((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 2) + 2) % 21) / 3) + ry_outer_outer) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) * 2) + 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 2) + 128) / 3) * 7)) + (ry_outer_out [...]
+ pad_temp_shared[((int)threadIdx.x)] = ((((1 <= ((((int)threadIdx.x) / 9) + ry_outer_outer)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 7)) + (ry_outer_outer * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 49)] = (((((1 <= ((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 49) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 98)] = (((((1 <= ((((((int)threadIdx.x) + 35) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 35) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 98) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 147)] = (((((1 <= ((((((int)threadIdx.x) + 21) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 21) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 147) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 196)] = (((((1 <= ((((((int)threadIdx.x) + 7) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 7) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 196) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 245)] = (((((1 <= ((((((int)threadIdx.x) + 56) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 56) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 245) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 294)] = (((((1 <= ((((((int)threadIdx.x) + 42) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 42) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 294) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 343)] = (((((1 <= ((((((int)threadIdx.x) + 28) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) + 28) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 343) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 392)] = ((((((((((int)threadIdx.x) + 14) % 63) / 9) + ry_outer_outer) < 8) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 392) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 441)] = ((((1 <= ((((int)threadIdx.x) / 9) + ry_outer_outer)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 7)) + (ry_outer_outer * 7)) + (((int)threadIdx.x) % 9)) + 335)] : 0.000000e+00f);
+ if (((int)threadIdx.x) < 14) {
+ pad_temp_shared[(((int)threadIdx.x) + 490)] = ((((((((((int)threadIdx.x) + 49) % 63) / 9) + ry_outer_outer) < 8) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 490) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
}
- if (((int)threadIdx.x) < 20) {
- pad_temp_shared[((((int)threadIdx.x) * 2) + 129)] = (((((1 <= (ry_outer_outer + ((((((int)threadIdx.x) * 2) / 3) + 1) % 7))) && ((ry_outer_outer + ((((((int)threadIdx.x) * 2) / 3) + 1) % 7)) < 8)) && (1 <= ((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)))) && (((((int)blockIdx.x) % 7) + ((((int)threadIdx.x) * 2) % 3)) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 2) / 3) * 7)) + (ry_outer_outer * 7)) + (((int)bloc [...]
+ kernel_shared[(((int)threadIdx.x) * 4)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) % 6) * 4) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 1)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) % 6) * 4) + 1) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 2)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) % 6) * 4) + 2) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 2) % 3))];
+ kernel_shared[((((int)threadIdx.x) * 4) + 3)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) / 3) + 1) & 7) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 196)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) + 4) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 197)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) + 5) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 2) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 198)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + (((((((int)threadIdx.x) * 4) / 3) + 2) & 7) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) * 4) % 3))];
+ }
+ if (((int)threadIdx.x) < 47) {
+ kernel_shared[((((int)threadIdx.x) * 4) + 199)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 49) / 6) * 4608)) + (rc_outer_outer * 72)) + ((((((((int)threadIdx.x) * 4) + 196) / 3) + 1) & 7) * 9)) + (ry_outer_outer * 3)) + (((((int)threadIdx.x) * 4) + 1) % 3))];
}
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 32)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 32) / 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) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 96)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 18432)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 160)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 160) / 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) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 224)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 224) / 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) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 288)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 55296)];
- kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 352)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 352) / 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) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 416)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 416) / 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) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 480)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 92160)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 544)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 544) / 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) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 608)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 608) / 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) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 672)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 129024)];
- kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 736)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 736) / 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) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 800)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 800) / 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) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 864)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 165888)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 928)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 928) / 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) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 992)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 992) / 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) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1056)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 202752)];
- kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1120)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1120) / 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) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 1184)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1184) / 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) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1248)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 239616)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1312)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1312) / 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) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((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) + 1376)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1376) / 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) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1440)] = kernel[((((((((((int)blockIdx.x) / 7) * 294912) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 276480)];
- kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((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) + 1504)] = kernel[(((((((((int)blockIdx.x) / 7) * 294912) + (((((int)threadIdx.x) + 1504) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
__syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[9] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[15] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[18] * kernel_shared[(((int)threadIdx.x) * 48)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
+ for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
+ for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[((rc_outer_inner * 12) + rx_outer_inner)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 192)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 216)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 3)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 195)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 219)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 6)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 198)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 222)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 9)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 201)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 225)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 48)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 240)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 72)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 264)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 51)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 243)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 75)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 267)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 54)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 246)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 78)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 270)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 57)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 249)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 81)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 273)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 96)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 288)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 120)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 312)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 99)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 291)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 123)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 315)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 102)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 294)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 126)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 318)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 105)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 297)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 129)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 321)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 144)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 336)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 168)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7))] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 360)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 147)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 339)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 171)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 363)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 150)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 342)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 174)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 126)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 366)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 153)]));
+ conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 345)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 177)]));
+ conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 252) + ((((int)threadIdx.x) / 7) * 9)) + rx_outer_inner) + (((int)threadIdx.x) % 7)) + 189)] * kernel_shared[(((rc_outer_inner * 12) + rx_outer_inner) + 369)]));
+ }
+ }
}
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[(((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7))] = max((conv2d_nchw[i1_inner] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 7)] = max((conv2d_nchw[(i1_inner + 2)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 14)] = max((conv2d_nchw[(i1_inner + 4)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 21)] = max((conv2d_nchw[(i1_inner + 6)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 28)] = max((conv2d_nchw[(i1_inner + 8)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 35)] = max((conv2d_nchw[(i1_inner + 10)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[((((((((int)blockIdx.x) / 7) * 3136) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + (((int)blockIdx.x) % 7)) + 42)] = max((conv2d_nchw[(i1_inner + 12)] + bias[((((((int)blockIdx.x) / 7) * 64) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i1_inner = 0; i1_inner < 8; ++i1_inner) {
+ compute[(((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x))] = max((conv2d_nchw[i1_inner] + bias[((((int)blockIdx.x) * 16) + i1_inner)]), 0.000000e+00f);
+ compute[((((((int)blockIdx.x) * 784) + (i1_inner * 49)) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + i1_inner) + 8)]), 0.000000e+00f);
}
}
</pre></div>
@@ -1559,7 +922,7 @@ 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> ( 2 minutes 22.015 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 8.042 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 e5a861ec9..8bf6b380f 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -876,7 +876,7 @@ 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)
- 9.6688 9.6681 9.7013 9.6371 0.0262
+ 9.8376 9.8295 9.9006 9.7827 0.0485
</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 ce6104a25..5f87e7d0c 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -895,7 +895,7 @@ 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)
- 773.9947 774.0448 778.1659 769.7735 3.4264
+ 764.8067 762.8136 771.0680 760.5385 4.5238
</pre></div>
</div>
</div>
@@ -917,7 +917,7 @@ 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 18.048 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 20.332 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 4bb9db3d5..863d85407 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -600,74 +600,30 @@ 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], []), placeholder_9: placeholder_16: Buffer(placeholder_14, float32, [128, 512], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_6: placeholder_19: Buffer(placeholder_11, float32, [4916, 16, 1], [])} {
- for (i0.outer.i1.outer.fused: int32, 0, 256) "parallel" {
- allocate(compute_4: Pointer(global float32), float32, [256]), storage_scope = global {
- for (i.inner.init: int32, 0, 16) {
- let cse_var_1: int32 = (i.inner.init*16)
- {
- compute_5: Buffer(compute_4, float32, [256], [])[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
+ preflattened_buffer_map = {placeholder_6: placeholder_15: Buffer(placeholder_11, float32, [4916, 16, 1], []), placeholder_5: placeholder_16: Buffer(placeholder_10, float32, [128, 256], []), placeholder_8: placeholder_17: Buffer(placeholder_13, int32, [33], []), placeholder_7: placeholder_18: Buffer(placeholder_12, int32, [4916], []), placeholder_9: placeholder_19: Buffer(placeholder_14, float32, [128, 512], []), compute_1: compute_3: Buffer(compute_2, float32, [128, 512], [])} {
+ for (i0.outer.i1.outer.fused: int32, 0, 64) "parallel" {
+ allocate(compute_4: Pointer(global float32), float32, [1024]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 32) {
+ for (j.init: int32, 0, 16) {
+ compute_5: Buffer(compute_4, float32, [1024], [])[(((i.outer.inner*512) + (i.inner.init*16)) + j.init)] = 0f32
+ }
}
- }
- for (elem_idx: int32, 0, let cse_var_2: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_3[(cse_var_2 + 1)] - placeholder_3[cse_var_2])) {
- for (i.inner: int32, 0, 16) {
- let cse_var_21: int32 = floormod(i0.outer.i1.outer.fused, 32)
- let cse_var_20: int32 = (i.inner*16)
- let cse_var_19: int32 = (elem_idx*16)
- let cse_var_18: int32 = (cse_var_20 + 10)
- 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 = (cse_var_20 + 1)
- let cse_var_3: int32 = ((floordiv(i0.outer.i1.outer.fused, 32)*4096) + (i.inner*256))
- {
- compute_5[cse_var_20] = (compute_5[cse_var_20] + (placeholder_1[((placeholder_3[cse_var_21]*16) + cse_var_19)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_4] = (compute_5[cse_var_4] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 1)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_12] = (compute_5[cse_var_12] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 2)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_11] = (compute_5[cse_var_11] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 3)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_10] = (compute_5[cse_var_10] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 4)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_9] = (compute_5[cse_var_9] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 5)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_8] = (compute_5[cse_var_8] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 6)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_7] = (compute_5[cse_var_7] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 7)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_6] = (compute_5[cse_var_6] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 8)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_5] = (compute_5[cse_var_5] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 9)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_18] = (compute_5[cse_var_18] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 10)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_17] = (compute_5[cse_var_17] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 11)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_16] = (compute_5[cse_var_16] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 12)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_15] = (compute_5[cse_var_15] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 13)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_14] = (compute_5[cse_var_14] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 14)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
- compute_5[cse_var_13] = (compute_5[cse_var_13] + (placeholder_1[(((placeholder_3[cse_var_21]*16) + cse_var_19) + 15)]*max(placeholder[(cse_var_3 + placeholder_2[(placeholder_3[cse_var_21] + elem_idx)])], 0f32)))
+ for (elem_idx: int32, 0, let cse_var_1: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_3[(cse_var_1 + 1)] - placeholder_3[cse_var_1])) {
+ for (i.inner: int32, 0, 32) {
+ for (j: int32, 0, 16) {
+ let cse_var_3: int32 = floormod(i0.outer.i1.outer.fused, 32)
+ let cse_var_2: int32 = (((i.outer.inner*512) + (i.inner*16)) + j)
+ compute_5[cse_var_2] = (compute_5[cse_var_2] + (placeholder_1[(((placeholder_3[cse_var_3]*16) + (elem_idx*16)) + j)]*max(placeholder[((((floordiv(i0.outer.i1.outer.fused, 32)*16384) + (i.outer.inner*8192)) + (i.inner*256)) + placeholder_2[(placeholder_3[cse_var_3] + elem_idx)])], 0f32)))
+ }
}
}
}
- for (i0.inner: int32, 0, 16) {
- let cse_var_22: int32 = (((floordiv(i0.outer.i1.outer.fused, 32)*8192) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16))
- compute[ramp(cse_var_22, 1, 16)] = max((compute_5[ramp((i0.inner*16), 1, 16)] + placeholder_4[ramp(cse_var_22, 1, 16)]), broadcast(0f32, 16))
+ for (i0.inner: int32, 0, 64) {
+ for (i1.inner: int32, 0, 16) {
+ let cse_var_4: int32 = ((((floordiv(i0.outer.i1.outer.fused, 32)*32768) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16)) + i1.inner)
+ compute[cse_var_4] = max((compute_5[((i0.inner*16) + i1.inner)] + placeholder_4[cse_var_4]), 0f32)
+ }
}
}
}
@@ -706,7 +662,7 @@ 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: 1.936 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.639 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 d708ec7a1..f0432cdec 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>00:44.450</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:46.013</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:43.547</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>00:00.236</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>00:00.223</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>00:00.223</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>00:00.220</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>00:45.154</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>00:00.233</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>00:00.221</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>00:00.203</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>00:00.201</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>
</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 890572b4a..dd16a66c4 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: 111.03/111.03 result: MeasureResult(costs=(0.0020850042083333333,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8102667331695557, timestamp=1652131803.5408685) [('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: 0.00/111.03 result: Traceback (most recent call last):
+No: 6 GFLOPS: 92.44/92.44 result: MeasureResult(costs=(0.00250425925,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9339966773986816, timestamp=1652162349.4977636) [('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: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 8 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 9 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 10 GFLOPS: 0.00/92.44 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: 0.00/111.03 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 11 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 12 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 13 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 14 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 15 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 16 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 17 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 18 GFLOPS: 0.00/92.44 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: 0.00/111.03 result: Traceback (most recent call last):
+No: 19 GFLOPS: 0.00/92.44 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: 0x00007f8cd9001fa2
+ 12: 0x00007f3309641fa2
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: 143.78/143.78 result: MeasureResult(costs=(0.0016100564603174603,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1321747303009033, timestamp=1652131829.0709817) [('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: 143.45/143.45 result: MeasureResult(costs=(0.0016138067200000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.3942267894744873, timestamp=1652162368.5986714) [('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: 0.001961
+Time cost of this operator: 0.001996
</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 07c20252a..c4d2f874f 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -553,10 +553,10 @@ the tuned operator.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>########## 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 314.6 98.727 (1, 2, 10, 10, 3) 2 1
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.134 0.984 (1, 6, 10, 10) 1 1
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.923 0.29 (1, 1, 10, 10, 3) 1 1
-Total_time - 318.657 - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.9 98.761 (1, 2, 10, 10, 3) 2 1
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.0 0.953 (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 - 314.801 - - - -
</pre></div>
</div>
</div>
@@ -608,10 +608,10 @@ 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 81.9 96.822 (1, 6, 10, 10, 1) 2 1
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.736 2.053 (1, 6, 10, 10) 1 1
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.952 1.125 (1, 1, 10, 10, 3) 1 1
-Total_time - 84.588 - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 79.35 96.737 (1, 6, 10, 10, 1) 2 1
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.77 2.158 (1, 6, 10, 10) 1 1
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.906 1.105 (1, 1, 10, 10, 3) 1 1
+Total_time - 82.026 - - - -
</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/micro_tflite.html b/docs/how_to/work_with_microtvm/micro_tflite.html
index 6ad26afd0..997a7fc13 100644
--- a/docs/how_to/work_with_microtvm/micro_tflite.html
+++ b/docs/how_to/work_with_microtvm/micro_tflite.html
@@ -622,6 +622,7 @@ TVM_DLL int32_t tvmgen_default_fused_nn_dense_add(void* args, int32_t* arg_type_
- ./codegen/host/src
- ./codegen/host/src/default_lib0.c
- ./codegen/host/src/default_lib1.c
+ - ./codegen/host/src/default_lib2.c
- ./executor-config
- ./executor-config/graph
- ./executor-config/graph/graph.json
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 246ebf3b2..d22598597 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -300,13 +300,13 @@
<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>00:43.296</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>00:48.573</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:39.264</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>00:03.445</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>00:00.199</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>00:00.196</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</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></li>
-<li><p><strong>00:00.192</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>
+<li><p><strong>00:44.401</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>00:03.621</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>00:00.193</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</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>00:00.176</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 f80814eac..d748c68a1 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>00:08.757</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:05.141</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:06.738</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>00:01.799</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>00:00.220</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>00:03.501</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>00:01.439</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>00:00.201</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 6ec69d9b3..32396ec43 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>00:05.700</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:05.327</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<ul class="simple">
-<li><p><strong>00:02.093</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>00:01.123</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>00:00.733</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>00:00.718</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>00:00.315</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>00:00.247</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>00:00.243</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>00:00.228</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>00:02.061</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>00:00.877</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>00:00.715</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>00:00.711</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>00:00.291</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>00:00.227</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>00:00.224</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>00:00.221</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>
</ul>
</div>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index dcf0ed2df..b613ff7d9 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/tmp1iib50ex/input0.cc'\nsource_filename = \"/tmp/tmp1iib50ex/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/tmp3c3njbhh/input0.cc'\nsource_filename = \"/tmp/tmp3c3njbhh/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/annotated.html b/docs/reference/api/doxygen/annotated.html
index 8f9c3ad0e..b2b0db4c8 100644
--- a/docs/reference/api/doxygen/annotated.html
+++ b/docs/reference/api/doxygen/annotated.html
@@ -547,13 +547,17 @@ $(function() {
<tr id="row_1_6_242_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structtvm_1_1relay_1_1YoloReorgAttrs.html" target="_self">YoloReorgAttrs</a></td><td class="desc">Attributes used in yolo reorg operators </td></tr>
<tr id="row_1_7_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_1_7_" class="arrow" onclick="toggleFolder('1_7_')">►</span><span class="icona"><span class="icon">N</span></span><a class="el" href="namespacetvm_1_1runtime.html" target="_self">runtime</a></td><td class="desc"></td></tr>
<tr id="row_1_7_0_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_1_7_0_" class="arrow" onclick="toggleFolder('1_7_0_')">►</span><span class="icona"><span class="icon">N</span></span><a class="el" href="namespacetvm_1_1runtime_1_1metadata.html" target="_self">metadata</a></td><td class="desc"></td></tr>
-<tr id="row_1_7_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" target="_self">ArrayAccessor</a></td><td class="desc">A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">A [...]
-<tr id="row_1_7_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html" target="_self">ArrayAccessor< const char *, ::tvm::runtime::String ></a></td><td class="desc">A specialization of <a class="el" href="classtvm_1_1runtime_1_1meta [...]
-<tr id="row_1_7_0_2_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html" target="_self">ArrayIterator</a></td><td class="desc">An iterator implementation that lazily instantiates the C++ wrapping Metadata class </td></tr>
-<tr id="row_1_7_0_3_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" target="_self">MetadataArray</a></td><td class="desc">Reference class for <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" title="Reference class for MetadataArray. ">MetadataArray</a> </td></tr>
-<tr id="row_1_7_0_4_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html" target="_self">MetadataArrayNode</a></td><td class="desc">Container for arrays in the metadata </td></tr>
-<tr id="row_1_7_0_5_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html" target="_self">MetadataBase</a></td><td class="desc">Reference class for the common <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> c [...]
-<tr id="row_1_7_0_6_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" target="_self">MetadataBaseNode</a></td><td class="desc">Common base class for all Metadata </td></tr>
+<tr id="row_1_7_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" target="_self">ArrayAccessor</a></td><td class="desc">A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">A [...]
+<tr id="row_1_7_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html" target="_self">ArrayAccessor< const char *, ::tvm::runtime::String ></a></td><td class="desc">A specialization of <a class="el" href="classtvm_1_1runtime_1_1meta [...]
+<tr id="row_1_7_0_2_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html" target="_self">ArrayIterator</a></td><td class="desc">An iterator implementation that lazily instantiates the C++ wrapping <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> class </td></tr>
+<tr id="row_1_7_0_3_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html" target="_self">Metadata</a></td><td class="desc"></td></tr>
+<tr id="row_1_7_0_4_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" target="_self">MetadataArray</a></td><td class="desc">Reference class for <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" title="Reference class for MetadataArray. ">MetadataArray</a> </td></tr>
+<tr id="row_1_7_0_5_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html" target="_self">MetadataArrayNode</a></td><td class="desc">Container for arrays in the metadata </td></tr>
+<tr id="row_1_7_0_6_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html" target="_self">MetadataBase</a></td><td class="desc">Reference class for the common <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> c [...]
+<tr id="row_1_7_0_7_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" target="_self">MetadataBaseNode</a></td><td class="desc">Common base class for all <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> </td></tr>
+<tr id="row_1_7_0_8_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataNode.html" target="_self">MetadataNode</a></td><td class="desc"></td></tr>
+<tr id="row_1_7_0_9_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1TensorInfo.html" target="_self">TensorInfo</a></td><td class="desc"></td></tr>
+<tr id="row_1_7_0_10_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1TensorInfoNode.html" target="_self">TensorInfoNode</a></td><td class="desc"></td></tr>
<tr id="row_1_7_1_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_1_7_1_" class="arrow" onclick="toggleFolder('1_7_1_')">►</span><span class="icona"><span class="icon">N</span></span><a class="el" href="namespacetvm_1_1runtime_1_1micro__rpc.html" target="_self">micro_rpc</a></td><td class="desc"></td></tr>
<tr id="row_1_7_1_0_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1FrameBuffer.html" target="_self">FrameBuffer</a></td><td class="desc"></td></tr>
<tr id="row_1_7_1_1_" class="even" style="display:none;"><td class="entry"><span style="width:64px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1Framer.html" target="_self">Framer</a></td><td class="desc"></td></tr>
@@ -1038,18 +1042,19 @@ $(function() {
<tr id="row_1_144_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structtvm_1_1WorkspaceMemoryPoolsNode.html" target="_self">WorkspaceMemoryPoolsNode</a></td><td class="desc"></td></tr>
<tr id="row_2_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structMemoryManagerInterface.html" target="_self">MemoryManagerInterface</a></td><td class="desc"></td></tr>
<tr id="row_3_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structtvm__workspace__t.html" target="_self">tvm_workspace_t</a></td><td class="desc"></td></tr>
-<tr id="row_4_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMArgs.html" target="_self">TVMArgs</a></td><td class="desc"></td></tr>
-<tr id="row_5_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMByteArray.html" target="_self">TVMByteArray</a></td><td class="desc">Byte array type used to pass in byte array When kTVMBytes is used as data type </td></tr>
-<tr id="row_6_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMFuncRegistry.html" target="_self">TVMFuncRegistry</a></td><td class="desc">A data structure that facilitates function lookup by C-string name </td></tr>
-<tr id="row_7_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMGraphExecutorGraphAttr.html" target="_self">TVMGraphExecutorGraphAttr</a></td><td class="desc"></td></tr>
-<tr id="row_8_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMMetadata.html" target="_self">TVMMetadata</a></td><td class="desc">Top-level metadata structure. Holds all other metadata types </td></tr>
-<tr id="row_9_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMModule.html" target="_self">TVMModule</a></td><td class="desc">Module container of TVM </td></tr>
-<tr id="row_10_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMMutableFuncRegistry.html" target="_self">TVMMutableFuncRegistry</a></td><td class="desc">A <a class="el" href="structTVMFuncRegistry.html" title="A data structure that facilitates function lookup by C-string name. ">TVMFuncRegistry</a> that supports adding and changing the functions </td></tr>
-<tr id="row_11_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMOpParam.html" target="_self">TVMOpParam</a></td><td class="desc">Operator attributes about tvm op </td></tr>
-<tr id="row_12_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMPackedFunc.html" target="_self">TVMPackedFunc</a></td><td class="desc"></td></tr>
-<tr id="row_13_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMParallelGroupEnv.html" target="_self">TVMParallelGroupEnv</a></td><td class="desc">Environment for TVM parallel task </td></tr>
-<tr id="row_14_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMTensorInfo.html" target="_self">TVMTensorInfo</a></td><td class="desc">Describes one tensor argument to <code>run_model</code>. NOTE: while TIR allows for other types of arguments, such as scalars, the AOT run_model function does not currently accept these. Therefore it's not possible to express thos [...]
-<tr id="row_15_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="unionTVMValue.html" target="_self">TVMValue</a></td><td class="desc">Union type of values being passed through API and function calls </td></tr>
+<tr id="row_4_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMAotExecutor.html" target="_self">TVMAotExecutor</a></td><td class="desc"></td></tr>
+<tr id="row_5_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMArgs.html" target="_self">TVMArgs</a></td><td class="desc"></td></tr>
+<tr id="row_6_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMByteArray.html" target="_self">TVMByteArray</a></td><td class="desc">Byte array type used to pass in byte array When kTVMBytes is used as data type </td></tr>
+<tr id="row_7_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMFuncRegistry.html" target="_self">TVMFuncRegistry</a></td><td class="desc">A data structure that facilitates function lookup by C-string name </td></tr>
+<tr id="row_8_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMGraphExecutorGraphAttr.html" target="_self">TVMGraphExecutorGraphAttr</a></td><td class="desc"></td></tr>
+<tr id="row_9_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMMetadata.html" target="_self">TVMMetadata</a></td><td class="desc">Top-level metadata structure. Holds all other metadata types </td></tr>
+<tr id="row_10_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMModule.html" target="_self">TVMModule</a></td><td class="desc">Module container of TVM </td></tr>
+<tr id="row_11_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMMutableFuncRegistry.html" target="_self">TVMMutableFuncRegistry</a></td><td class="desc">A <a class="el" href="structTVMFuncRegistry.html" title="A data structure that facilitates function lookup by C-string name. ">TVMFuncRegistry</a> that supports adding and changing the functions </td></tr>
+<tr id="row_12_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMOpParam.html" target="_self">TVMOpParam</a></td><td class="desc">Operator attributes about tvm op </td></tr>
+<tr id="row_13_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMPackedFunc.html" target="_self">TVMPackedFunc</a></td><td class="desc"></td></tr>
+<tr id="row_14_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMParallelGroupEnv.html" target="_self">TVMParallelGroupEnv</a></td><td class="desc">Environment for TVM parallel task </td></tr>
+<tr id="row_15_"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="structTVMTensorInfo.html" target="_self">TVMTensorInfo</a></td><td class="desc">Describes one tensor argument to <code>run_model</code>. NOTE: while TIR allows for other types of arguments, such as scalars, the AOT run_model function does not currently accept these. Therefore it's not possible to express those in this met [...]
+<tr id="row_16_" class="even"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="unionTVMValue.html" target="_self">TVMValue</a></td><td class="desc">Union type of values being passed through API and function calls </td></tr>
</table>
</div><!-- directory -->
</div><!-- contents -->
diff --git a/docs/reference/api/doxygen/aot__executor_8h.html b/docs/reference/api/doxygen/aot__executor_8h.html
new file mode 100644
index 000000000..621fa52a5
--- /dev/null
+++ b/docs/reference/api/doxygen/aot__executor_8h.html
@@ -0,0 +1,370 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml">
+<head>
+<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
+<meta http-equiv="X-UA-Compatible" content="IE=9"/>
+<meta name="generator" content="Doxygen 1.8.13"/>
+<meta name="viewport" content="width=device-width, initial-scale=1"/>
+<title>tvm: include/tvm/runtime/crt/aot_executor.h File Reference</title>
+<link href="tabs.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="jquery.js"></script>
+<script type="text/javascript" src="dynsections.js"></script>
+<link href="search/search.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="search/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+</head>
+<body>
+<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
+<div id="titlearea">
+<table cellspacing="0" cellpadding="0">
+ <tbody>
+ <tr style="height: 56px;">
+ <td id="projectalign" style="padding-left: 0.5em;">
+ <div id="projectname">tvm
+ </div>
+ </td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.13 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+<script type="text/javascript" src="menudata.js"></script>
+<script type="text/javascript" src="menu.js"></script>
+<script type="text/javascript">
+$(function() {
+ initMenu('',true,false,'search.php','Search');
+ $(document).ready(function() { init_search(); });
+});
+</script>
+<div id="main-nav"></div>
+<!-- window showing the filter options -->
+<div id="MSearchSelectWindow"
+ onmouseover="return searchBox.OnSearchSelectShow()"
+ onmouseout="return searchBox.OnSearchSelectHide()"
+ onkeydown="return searchBox.OnSearchSelectKey(event)">
+</div>
+
+<!-- iframe showing the search results (closed by default) -->
+<div id="MSearchResultsWindow">
+<iframe src="javascript:void(0)" frameborder="0"
+ name="MSearchResults" id="MSearchResults">
+</iframe>
+</div>
+
+<div id="nav-path" class="navpath">
+ <ul>
+<li class="navelem"><a class="el" href="dir_d44c64559bbebec7f509842c48db8b23.html">include</a></li><li class="navelem"><a class="el" href="dir_b4c7d8e826c599ba55146c099a14beb5.html">tvm</a></li><li class="navelem"><a class="el" href="dir_404558507ed35459f0d7a6d81d8c508d.html">runtime</a></li><li class="navelem"><a class="el" href="dir_a2900df4deca8dd2bcded616f0fe650a.html">crt</a></li> </ul>
+</div>
+</div><!-- top -->
+<div class="header">
+ <div class="summary">
+<a href="#nested-classes">Classes</a> |
+<a href="#typedef-members">Typedefs</a> |
+<a href="#func-members">Functions</a> </div>
+ <div class="headertitle">
+<div class="title">aot_executor.h File Reference</div> </div>
+</div><!--header-->
+<div class="contents">
+
+<p>AoT Executor.
+<a href="#details">More...</a></p>
+<div class="textblock"><code>#include <dlpack/dlpack.h></code><br />
+<code>#include <tvm/runtime/crt/internal/common/ndarray.h></code><br />
+<code>#include <<a class="el" href="metadata__types_8h_source.html">tvm/runtime/metadata_types.h</a>></code><br />
+</div><div class="textblock"><div class="dynheader">
+Include dependency graph for aot_executor.h:</div>
+<div class="dyncontent">
+<div class="center"><iframe scrolling="no" frameborder="0" src="aot__executor_8h__incl.svg" width="407" height="306"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
+</div>
+</div>
+</div>
+<p><a href="aot__executor_8h_source.html">Go to the source code of this file.</a></p>
+<table class="memberdecls">
+<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="nested-classes"></a>
+Classes</h2></td></tr>
+<tr class="memitem:"><td class="memItemLeft" align="right" valign="top">struct  </td><td class="memItemRight" valign="bottom"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+</table><table class="memberdecls">
+<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="typedef-members"></a>
+Typedefs</h2></td></tr>
+<tr class="memitem:a4599d4a62bfcf460b9682d8a715360c7"><td class="memItemLeft" align="right" valign="top">typedef struct <a class="el" href="structTVMMetadata.html">TVMMetadata</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#a4599d4a62bfcf460b9682d8a715360c7">TVMMetadata</a></td></tr>
+<tr class="separator:a4599d4a62bfcf460b9682d8a715360c7"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:ad090188290895a09067753e79d57077b"><td class="memItemLeft" align="right" valign="top">typedef struct <a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#ad090188290895a09067753e79d57077b">TVMAotExecutor</a></td></tr>
+<tr class="separator:ad090188290895a09067753e79d57077b"><td class="memSeparator" colspan="2"> </td></tr>
+</table><table class="memberdecls">
+<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="func-members"></a>
+Functions</h2></td></tr>
+<tr class="memitem:ab663a086fc2c4f6df166625a9ac83ee1"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1">TVMAotExecutor_Create</a> (<a class="el" href="c__runtime__api_8h.html#ac4b0321c489287dcf4637e15328298ce">TVMModuleHandle</a> module_handle, const DLDevice device, <a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> **executor, const char *m [...]
+<tr class="memdesc:ab663a086fc2c4f6df166625a9ac83ee1"><td class="mdescLeft"> </td><td class="mdescRight">Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. <a href="#ab663a086fc2c4f6df166625a9ac83ee1">More...</a><br /></td></tr>
+<tr class="separator:ab663a086fc2c4f6df166625a9ac83ee1"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a15512a380d3f9cbd9c24ad9289f1dfd6"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#a15512a380d3f9cbd9c24ad9289f1dfd6">TVMAotExecutor_Release</a> (<a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> *executor, const DLDevice device)</td></tr>
+<tr class="memdesc:a15512a380d3f9cbd9c24ad9289f1dfd6"><td class="mdescLeft"> </td><td class="mdescRight">Release the AoT executor created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. <a href="#a15512a380d3f9cbd9c24ad9289f1dfd6">More...</a><br /></td></tr>
+<tr class="separator:a15512a380d3f9cbd9c24ad9289f1dfd6"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a84924c70598933ebb941961f77cdd977"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#a84924c70598933ebb941961f77cdd977">TVMAotExecutor_GetNumInputs</a> (<a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> *executor)</td></tr>
+<tr class="memdesc:a84924c70598933ebb941961f77cdd977"><td class="mdescLeft"> </td><td class="mdescRight">Return the number of inputs. <a href="#a84924c70598933ebb941961f77cdd977">More...</a><br /></td></tr>
+<tr class="separator:a84924c70598933ebb941961f77cdd977"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a7f42e0f9479896def33b3427f9a97518"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#a7f42e0f9479896def33b3427f9a97518">TVMAotExecutor_GetNumOutputs</a> (<a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> *executor)</td></tr>
+<tr class="memdesc:a7f42e0f9479896def33b3427f9a97518"><td class="mdescLeft"> </td><td class="mdescRight">Return the number of outputs. <a href="#a7f42e0f9479896def33b3427f9a97518">More...</a><br /></td></tr>
+<tr class="separator:a7f42e0f9479896def33b3427f9a97518"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:ad3b4f9b4d0b73d9c1b7e42d77076ffdb"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#ad3b4f9b4d0b73d9c1b7e42d77076ffdb">TVMAotExecutor_GetInputIndex</a> (<a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> *executor, const char *name)</td></tr>
+<tr class="memdesc:ad3b4f9b4d0b73d9c1b7e42d77076ffdb"><td class="mdescLeft"> </td><td class="mdescRight">Return the input index of the specified input name. <a href="#ad3b4f9b4d0b73d9c1b7e42d77076ffdb">More...</a><br /></td></tr>
+<tr class="separator:ad3b4f9b4d0b73d9c1b7e42d77076ffdb"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:aa576382ee317ceef3e1858d0a953df95"><td class="memItemLeft" align="right" valign="top">int </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor_8h.html#aa576382ee317ceef3e1858d0a953df95">TVMAotExecutor_Run</a> (<a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> *executor)</td></tr>
+<tr class="memdesc:aa576382ee317ceef3e1858d0a953df95"><td class="mdescLeft"> </td><td class="mdescRight">Run the generated program. <a href="#aa576382ee317ceef3e1858d0a953df95">More...</a><br /></td></tr>
+<tr class="separator:aa576382ee317ceef3e1858d0a953df95"><td class="memSeparator" colspan="2"> </td></tr>
+</table>
+<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
+<div class="textblock"><p>AoT Executor. </p>
+</div><h2 class="groupheader">Typedef Documentation</h2>
+<a id="ad090188290895a09067753e79d57077b"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#ad090188290895a09067753e79d57077b">◆ </a></span>TVMAotExecutor</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">typedef struct <a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> <a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+</div>
+</div>
+<a id="a4599d4a62bfcf460b9682d8a715360c7"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a4599d4a62bfcf460b9682d8a715360c7">◆ </a></span>TVMMetadata</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">typedef struct <a class="el" href="structTVMMetadata.html">TVMMetadata</a> <a class="el" href="structTVMMetadata.html">TVMMetadata</a></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+</div>
+</div>
+<h2 class="groupheader">Function Documentation</h2>
+<a id="ab663a086fc2c4f6df166625a9ac83ee1"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#ab663a086fc2c4f6df166625a9ac83ee1">◆ </a></span>TVMAotExecutor_Create()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_Create </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="c__runtime__api_8h.html#ac4b0321c489287dcf4637e15328298ce">TVMModuleHandle</a> </td>
+ <td class="paramname"><em>module_handle</em>, </td>
+ </tr>
+ <tr>
+ <td class="paramkey"></td>
+ <td></td>
+ <td class="paramtype">const DLDevice </td>
+ <td class="paramname"><em>device</em>, </td>
+ </tr>
+ <tr>
+ <td class="paramkey"></td>
+ <td></td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> ** </td>
+ <td class="paramname"><em>executor</em>, </td>
+ </tr>
+ <tr>
+ <td class="paramkey"></td>
+ <td></td>
+ <td class="paramtype">const char * </td>
+ <td class="paramname"><em>module_name</em> </td>
+ </tr>
+ <tr>
+ <td></td>
+ <td>)</td>
+ <td></td><td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">module_handle</td><td>TVM Module that exposes the functions to call. </td></tr>
+ <tr><td class="paramname">device</td><td>Runtime execution device, only supports device type kDLCPU, index 0. </td></tr>
+ <tr><td class="paramname">executor</td><td>Pointer which receives a pointer to the newly-created instance. </td></tr>
+ <tr><td class="paramname">module_name</td><td>TVM Module name prefix, typically "default". </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>0 if successful. </dd></dl>
+
+</div>
+</div>
+<a id="ad3b4f9b4d0b73d9c1b7e42d77076ffdb"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#ad3b4f9b4d0b73d9c1b7e42d77076ffdb">◆ </a></span>TVMAotExecutor_GetInputIndex()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_GetInputIndex </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> * </td>
+ <td class="paramname"><em>executor</em>, </td>
+ </tr>
+ <tr>
+ <td class="paramkey"></td>
+ <td></td>
+ <td class="paramtype">const char * </td>
+ <td class="paramname"><em>name</em> </td>
+ </tr>
+ <tr>
+ <td></td>
+ <td>)</td>
+ <td></td><td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Return the input index of the specified input name. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">executor</td><td>Pointer to executor instance, created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </td></tr>
+ <tr><td class="paramname">name</td><td>Input name for retrieving index. </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>Input index. </dd></dl>
+
+</div>
+</div>
+<a id="a84924c70598933ebb941961f77cdd977"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a84924c70598933ebb941961f77cdd977">◆ </a></span>TVMAotExecutor_GetNumInputs()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_GetNumInputs </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> * </td>
+ <td class="paramname"><em>executor</em></td><td>)</td>
+ <td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Return the number of inputs. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">executor</td><td>Pointer to executor instance, created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>Number of inputs. </dd></dl>
+
+</div>
+</div>
+<a id="a7f42e0f9479896def33b3427f9a97518"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a7f42e0f9479896def33b3427f9a97518">◆ </a></span>TVMAotExecutor_GetNumOutputs()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_GetNumOutputs </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> * </td>
+ <td class="paramname"><em>executor</em></td><td>)</td>
+ <td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Return the number of outputs. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">executor</td><td>Pointer to executor instance, created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>Number of outputs. </dd></dl>
+
+</div>
+</div>
+<a id="a15512a380d3f9cbd9c24ad9289f1dfd6"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a15512a380d3f9cbd9c24ad9289f1dfd6">◆ </a></span>TVMAotExecutor_Release()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_Release </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> * </td>
+ <td class="paramname"><em>executor</em>, </td>
+ </tr>
+ <tr>
+ <td class="paramkey"></td>
+ <td></td>
+ <td class="paramtype">const DLDevice </td>
+ <td class="paramname"><em>device</em> </td>
+ </tr>
+ <tr>
+ <td></td>
+ <td>)</td>
+ <td></td><td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Release the AoT executor created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">executor</td><td>Pointer to executor instance, created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </td></tr>
+ <tr><td class="paramname">device</td><td>Runtime execution device, only supports device type kDLCPU, index 0. </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>0 if successful. </dd></dl>
+
+</div>
+</div>
+<a id="aa576382ee317ceef3e1858d0a953df95"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#aa576382ee317ceef3e1858d0a953df95">◆ </a></span>TVMAotExecutor_Run()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">int TVMAotExecutor_Run </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="structTVMAotExecutor.html">TVMAotExecutor</a> * </td>
+ <td class="paramname"><em>executor</em></td><td>)</td>
+ <td></td>
+ </tr>
+ </table>
+</div><div class="memdoc">
+
+<p>Run the generated program. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+ <table class="params">
+ <tr><td class="paramname">executor</td><td>Pointer to executor instance, created by <a class="el" href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1" title="Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. ">TVMAotExecutor_Create()</a>. </td></tr>
+ </table>
+ </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>0 if successful. </dd></dl>
+
+</div>
+</div>
+</div><!-- contents -->
+<!-- start footer part -->
+<hr class="footer"/><address class="footer"><small>
+Generated by  <a href="http://www.doxygen.org/index.html">
+<img class="footer" src="doxygen.png" alt="doxygen"/>
+</a> 1.8.13
+</small></address>
+</body>
+</html>
diff --git a/docs/reference/api/doxygen/aot__executor_8h__incl.svg b/docs/reference/api/doxygen/aot__executor_8h__incl.svg
new file mode 100644
index 000000000..f66cbd763
--- /dev/null
+++ b/docs/reference/api/doxygen/aot__executor_8h__incl.svg
@@ -0,0 +1,119 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.40.1 (20161225.0304)
+ -->
+<!-- Title: include/tvm/runtime/crt/aot_executor.h Pages: 1 -->
+<svg width="305pt" height="229pt"
+ viewBox="0.00 0.00 304.79 229.00" 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(4 225)">
+<title>include/tvm/runtime/crt/aot_executor.h</title>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-225 300.7862,-225 300.7862,4 -4,4"/>
+<!-- Node0 -->
+<g id="node1" class="node">
+<title>Node0</title>
+<polygon fill="#bfbfbf" stroke="#000000" points="25.2862,-190.5 25.2862,-220.5 141.2862,-220.5 141.2862,-190.5 25.2862,-190.5"/>
+<text text-anchor="start" x="33.2862" y="-208.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="83.2862" y="-197.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/aot_executor.h</text>
+</g>
+<!-- Node1 -->
+<g id="node2" class="node">
+<title>Node1</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="16.7862,-.5 16.7862,-19.5 109.7862,-19.5 109.7862,-.5 16.7862,-.5"/>
+<text text-anchor="middle" x="63.2862" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">dlpack/dlpack.h</text>
+</g>
+<!-- Node0->Node1 -->
+<g id="edge1" class="edge">
+<title>Node0->Node1</title>
+<path fill="none" stroke="#191970" d="M46.9345,-190.3589C31.8014,-182.0205 15.7992,-170.0521 7.2862,-154 -15.8344,-110.4039 23.1951,-55.1012 46.9891,-27.4314"/>
+<polygon fill="#191970" stroke="#191970" points="49.7682,-29.5732 53.7987,-19.775 44.5377,-24.9212 49.7682,-29.5732"/>
+</g>
+<!-- Node2 -->
+<g id="node3" class="node">
+<title>Node2</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="16.2862,-123.5 16.2862,-153.5 150.2862,-153.5 150.2862,-123.5 16.2862,-123.5"/>
+<text text-anchor="start" x="24.2862" y="-141.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/crt/internal</text>
+<text text-anchor="middle" x="83.2862" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/common/ndarray.h</text>
+</g>
+<!-- Node0->Node2 -->
+<g id="edge2" class="edge">
+<title>Node0->Node2</title>
+<path fill="none" stroke="#191970" d="M83.2862,-190.2967C83.2862,-182.5013 83.2862,-172.7991 83.2862,-163.9064"/>
+<polygon fill="#191970" stroke="#191970" points="86.7863,-163.6431 83.2862,-153.6432 79.7863,-163.6432 86.7863,-163.6431"/>
+</g>
+<!-- Node3 -->
+<g id="node4" class="node">
+<title>Node3</title>
+<g id="a_node4"><a xlink:href="metadata__types_8h.html" target="_top" xlink:title="Defines types which can be used in metadata here which are also shared between C and C++ code bases...">
+<polygon fill="#ffffff" stroke="#000000" points="168.2862,-123.5 168.2862,-153.5 296.2862,-153.5 296.2862,-123.5 168.2862,-123.5"/>
+<text text-anchor="start" x="176.2862" y="-141.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/metadata</text>
+<text text-anchor="middle" x="232.2862" y="-130.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_types.h</text>
+</a>
+</g>
+</g>
+<!-- Node0->Node3 -->
+<g id="edge3" class="edge">
+<title>Node0->Node3</title>
+<path fill="none" stroke="#191970" d="M116.7247,-190.4639C138.2169,-180.7996 166.3393,-168.154 189.4604,-157.7572"/>
+<polygon fill="#191970" stroke="#191970" points="191.0169,-160.895 198.7018,-153.6017 188.1461,-154.5107 191.0169,-160.895"/>
+</g>
+<!-- Node4 -->
+<g id="node5" class="node">
+<title>Node4</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="84.7862,-62 84.7862,-81 149.7862,-81 149.7862,-62 84.7862,-62"/>
+<text text-anchor="middle" x="117.2862" y="-69" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">inttypes.h</text>
+</g>
+<!-- Node3->Node4 -->
+<g id="edge4" class="edge">
+<title>Node3->Node4</title>
+<path fill="none" stroke="#191970" d="M206.4779,-123.4639C187.3562,-112.3234 161.4345,-97.2212 142.4644,-86.1691"/>
+<polygon fill="#191970" stroke="#191970" points="144.173,-83.1138 133.7705,-81.1039 140.6491,-89.1622 144.173,-83.1138"/>
+</g>
+<!-- Node5 -->
+<g id="node6" class="node">
+<title>Node5</title>
+<g id="a_node6"><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="167.7862,-56.5 167.7862,-86.5 296.7862,-86.5 296.7862,-56.5 167.7862,-56.5"/>
+<text text-anchor="start" x="175.7862" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/c_runtime</text>
+<text text-anchor="middle" x="232.2862" y="-63.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_api.h</text>
+</a>
+</g>
+</g>
+<!-- Node3->Node5 -->
+<g id="edge5" class="edge">
+<title>Node3->Node5</title>
+<path fill="none" stroke="#191970" d="M232.2862,-123.2967C232.2862,-115.5013 232.2862,-105.7991 232.2862,-96.9064"/>
+<polygon fill="#191970" stroke="#191970" points="235.7863,-96.6431 232.2862,-86.6432 228.7863,-96.6432 235.7863,-96.6431"/>
+</g>
+<!-- Node5->Node1 -->
+<g id="edge6" class="edge">
+<title>Node5->Node1</title>
+<path fill="none" stroke="#191970" d="M190.9441,-56.4554C162.746,-46.194 125.7831,-32.7429 98.8919,-22.9571"/>
+<polygon fill="#191970" stroke="#191970" points="100.0692,-19.6611 89.4752,-19.5303 97.6754,-26.239 100.0692,-19.6611"/>
+</g>
+<!-- Node6 -->
+<g id="node7" class="node">
+<title>Node6</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="167.2862,-.5 167.2862,-19.5 223.2862,-19.5 223.2862,-.5 167.2862,-.5"/>
+<text text-anchor="middle" x="195.2862" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stddef.h</text>
+</g>
+<!-- Node5->Node6 -->
+<g id="edge7" class="edge">
+<title>Node5->Node6</title>
+<path fill="none" stroke="#191970" d="M223.1401,-56.2977C218.1111,-47.9388 211.8307,-37.4997 206.5071,-28.6509"/>
+<polygon fill="#191970" stroke="#191970" points="209.3647,-26.6114 201.2103,-19.8469 203.3665,-30.2201 209.3647,-26.6114"/>
+</g>
+<!-- Node7 -->
+<g id="node8" class="node">
+<title>Node7</title>
+<polygon fill="#ffffff" stroke="#bfbfbf" points="241.7862,-.5 241.7862,-19.5 294.7862,-19.5 294.7862,-.5 241.7862,-.5"/>
+<text text-anchor="middle" x="268.2862" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">stdint.h</text>
+</g>
+<!-- Node5->Node7 -->
+<g id="edge8" class="edge">
+<title>Node5->Node7</title>
+<path fill="none" stroke="#191970" d="M241.1851,-56.2977C246.0781,-47.9388 252.1888,-37.4997 257.3686,-28.6509"/>
+<polygon fill="#191970" stroke="#191970" points="260.4908,-30.2452 262.5221,-19.8469 254.4497,-26.7089 260.4908,-30.2452"/>
+</g>
+</g>
+</svg>
diff --git a/docs/reference/api/doxygen/aot__executor_8h_source.html b/docs/reference/api/doxygen/aot__executor_8h_source.html
new file mode 100644
index 000000000..c51e5adb3
--- /dev/null
+++ b/docs/reference/api/doxygen/aot__executor_8h_source.html
@@ -0,0 +1,93 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml">
+<head>
+<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
+<meta http-equiv="X-UA-Compatible" content="IE=9"/>
+<meta name="generator" content="Doxygen 1.8.13"/>
+<meta name="viewport" content="width=device-width, initial-scale=1"/>
+<title>tvm: include/tvm/runtime/crt/aot_executor.h Source File</title>
+<link href="tabs.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="jquery.js"></script>
+<script type="text/javascript" src="dynsections.js"></script>
+<link href="search/search.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="search/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+</head>
+<body>
+<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
+<div id="titlearea">
+<table cellspacing="0" cellpadding="0">
+ <tbody>
+ <tr style="height: 56px;">
+ <td id="projectalign" style="padding-left: 0.5em;">
+ <div id="projectname">tvm
+ </div>
+ </td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.13 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+<script type="text/javascript" src="menudata.js"></script>
+<script type="text/javascript" src="menu.js"></script>
+<script type="text/javascript">
+$(function() {
+ initMenu('',true,false,'search.php','Search');
+ $(document).ready(function() { init_search(); });
+});
+</script>
+<div id="main-nav"></div>
+<!-- window showing the filter options -->
+<div id="MSearchSelectWindow"
+ onmouseover="return searchBox.OnSearchSelectShow()"
+ onmouseout="return searchBox.OnSearchSelectHide()"
+ onkeydown="return searchBox.OnSearchSelectKey(event)">
+</div>
+
+<!-- iframe showing the search results (closed by default) -->
+<div id="MSearchResultsWindow">
+<iframe src="javascript:void(0)" frameborder="0"
+ name="MSearchResults" id="MSearchResults">
+</iframe>
+</div>
+
+<div id="nav-path" class="navpath">
+ <ul>
+<li class="navelem"><a class="el" href="dir_d44c64559bbebec7f509842c48db8b23.html">include</a></li><li class="navelem"><a class="el" href="dir_b4c7d8e826c599ba55146c099a14beb5.html">tvm</a></li><li class="navelem"><a class="el" href="dir_404558507ed35459f0d7a6d81d8c508d.html">runtime</a></li><li class="navelem"><a class="el" href="dir_a2900df4deca8dd2bcded616f0fe650a.html">crt</a></li> </ul>
+</div>
+</div><!-- top -->
+<div class="header">
+ <div class="headertitle">
+<div class="title">aot_executor.h</div> </div>
+</div><!--header-->
+<div class="contents">
+<a href="aot__executor_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">/*</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment"> * Licensed to the Apache Software Foundation (ASF) under one</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> * or mo [...]
+<div class="ttc" id="c__runtime__api_8h_html_ac4b0321c489287dcf4637e15328298ce"><div class="ttname"><a href="c__runtime__api_8h.html#ac4b0321c489287dcf4637e15328298ce">TVMModuleHandle</a></div><div class="ttdeci">void * TVMModuleHandle</div><div class="ttdoc">Handle to TVM runtime modules. </div><div class="ttdef"><b>Definition:</b> c_runtime_api.h:163</div></div>
+<div class="ttc" id="aot__executor_8h_html_a84924c70598933ebb941961f77cdd977"><div class="ttname"><a href="aot__executor_8h.html#a84924c70598933ebb941961f77cdd977">TVMAotExecutor_GetNumInputs</a></div><div class="ttdeci">int TVMAotExecutor_GetNumInputs(TVMAotExecutor *executor)</div><div class="ttdoc">Return the number of inputs. </div></div>
+<div class="ttc" id="structTVMAotExecutor_html_a524b567769ce3140e8f0235369432e76"><div class="ttname"><a href="structTVMAotExecutor.html#a524b567769ce3140e8f0235369432e76">TVMAotExecutor::metadata</a></div><div class="ttdeci">const TVMMetadata * metadata</div><div class="ttdoc">The top-level metadata structure supplied by the generated code. </div><div class="ttdef"><b>Definition:</b> aot_executor.h:39</div></div>
+<div class="ttc" id="aot__executor_8h_html_ad090188290895a09067753e79d57077b"><div class="ttname"><a href="aot__executor_8h.html#ad090188290895a09067753e79d57077b">TVMAotExecutor</a></div><div class="ttdeci">struct TVMAotExecutor TVMAotExecutor</div></div>
+<div class="ttc" id="structTVMAotExecutor_html_a83eeec227561f72783f1dbfecd814d61"><div class="ttname"><a href="structTVMAotExecutor.html#a83eeec227561f72783f1dbfecd814d61">TVMAotExecutor::num_args</a></div><div class="ttdeci">int64_t num_args</div><div class="ttdef"><b>Definition:</b> aot_executor.h:46</div></div>
+<div class="ttc" id="structTVMAotExecutor_html"><div class="ttname"><a href="structTVMAotExecutor.html">TVMAotExecutor</a></div><div class="ttdef"><b>Definition:</b> aot_executor.h:37</div></div>
+<div class="ttc" id="aot__executor_8h_html_a15512a380d3f9cbd9c24ad9289f1dfd6"><div class="ttname"><a href="aot__executor_8h.html#a15512a380d3f9cbd9c24ad9289f1dfd6">TVMAotExecutor_Release</a></div><div class="ttdeci">int TVMAotExecutor_Release(TVMAotExecutor *executor, const DLDevice device)</div><div class="ttdoc">Release the AoT executor created by TVMAotExecutor_Create(). </div></div>
+<div class="ttc" id="aot__executor_8h_html_ad3b4f9b4d0b73d9c1b7e42d77076ffdb"><div class="ttname"><a href="aot__executor_8h.html#ad3b4f9b4d0b73d9c1b7e42d77076ffdb">TVMAotExecutor_GetInputIndex</a></div><div class="ttdeci">int TVMAotExecutor_GetInputIndex(TVMAotExecutor *executor, const char *name)</div><div class="ttdoc">Return the input index of the specified input name. </div></div>
+<div class="ttc" id="structTVMAotExecutor_html_a556d142321ffa158b0b9190e84d6a086"><div class="ttname"><a href="structTVMAotExecutor.html#a556d142321ffa158b0b9190e84d6a086">TVMAotExecutor::device</a></div><div class="ttdeci">DLDevice device</div><div class="ttdoc">The device type. </div><div class="ttdef"><b>Definition:</b> aot_executor.h:43</div></div>
+<div class="ttc" id="structTVMAotExecutor_html_a0d4158663d39f79d88d2bc0355c9f1eb"><div class="ttname"><a href="structTVMAotExecutor.html#a0d4158663d39f79d88d2bc0355c9f1eb">TVMAotExecutor::module_handle</a></div><div class="ttdeci">TVMModuleHandle module_handle</div><div class="ttdoc">The code module that contains the compiled model. </div><div class="ttdef"><b>Definition:</b> aot_executor.h:41</div></div>
+<div class="ttc" id="structTVMAotExecutor_html_abfd0f6fd9189e652fa7a89d946db4840"><div class="ttname"><a href="structTVMAotExecutor.html#abfd0f6fd9189e652fa7a89d946db4840">TVMAotExecutor::args</a></div><div class="ttdeci">TVMNDArray * args</div><div class="ttdoc">List of allocated arguments, input(s), output(s), and pool(s) </div><div class="ttdef"><b>Definition:</b> aot_executor.h:45</div></div>
+<div class="ttc" id="aot__executor_8h_html_ab663a086fc2c4f6df166625a9ac83ee1"><div class="ttname"><a href="aot__executor_8h.html#ab663a086fc2c4f6df166625a9ac83ee1">TVMAotExecutor_Create</a></div><div class="ttdeci">int TVMAotExecutor_Create(TVMModuleHandle module_handle, const DLDevice device, TVMAotExecutor **executor, const char *module_name)</div><div class="ttdoc">Allocate a new AotExecutor with TVMPlatformMemoryAllocate and initialize it. </div></div>
+<div class="ttc" id="metadata__types_8h_html"><div class="ttname"><a href="metadata__types_8h.html">metadata_types.h</a></div><div class="ttdoc">Defines types which can be used in metadata here which are also shared between C and C++ code bases...</div></div>
+<div class="ttc" id="aot__executor_8h_html_a7f42e0f9479896def33b3427f9a97518"><div class="ttname"><a href="aot__executor_8h.html#a7f42e0f9479896def33b3427f9a97518">TVMAotExecutor_GetNumOutputs</a></div><div class="ttdeci">int TVMAotExecutor_GetNumOutputs(TVMAotExecutor *executor)</div><div class="ttdoc">Return the number of outputs. </div></div>
+<div class="ttc" id="aot__executor_8h_html_aa576382ee317ceef3e1858d0a953df95"><div class="ttname"><a href="aot__executor_8h.html#aa576382ee317ceef3e1858d0a953df95">TVMAotExecutor_Run</a></div><div class="ttdeci">int TVMAotExecutor_Run(TVMAotExecutor *executor)</div><div class="ttdoc">Run the generated program. </div></div>
+</div><!-- fragment --></div><!-- contents -->
+<!-- start footer part -->
+<hr class="footer"/><address class="footer"><small>
+Generated by  <a href="http://www.doxygen.org/index.html">
+<img class="footer" src="doxygen.png" alt="doxygen"/>
+</a> 1.8.13
+</small></address>
+</body>
+</html>
diff --git a/docs/reference/api/doxygen/graph__executor__module_8h.html b/docs/reference/api/doxygen/aot__executor__module_8h.html
similarity index 77%
copy from docs/reference/api/doxygen/graph__executor__module_8h.html
copy to docs/reference/api/doxygen/aot__executor__module_8h.html
index 93a251a3b..ebb7d44c0 100644
--- a/docs/reference/api/doxygen/graph__executor__module_8h.html
+++ b/docs/reference/api/doxygen/aot__executor__module_8h.html
@@ -5,7 +5,7 @@
<meta http-equiv="X-UA-Compatible" content="IE=9"/>
<meta name="generator" content="Doxygen 1.8.13"/>
<meta name="viewport" content="width=device-width, initial-scale=1"/>
-<title>tvm: include/tvm/runtime/crt/graph_executor_module.h File Reference</title>
+<title>tvm: include/tvm/runtime/crt/aot_executor_module.h File Reference</title>
<link href="tabs.css" rel="stylesheet" type="text/css"/>
<script type="text/javascript" src="jquery.js"></script>
<script type="text/javascript" src="dynsections.js"></script>
@@ -65,34 +65,34 @@ $(function() {
<div class="summary">
<a href="#func-members">Functions</a> </div>
<div class="headertitle">
-<div class="title">graph_executor_module.h File Reference</div> </div>
+<div class="title">aot_executor_module.h File Reference</div> </div>
</div><!--header-->
<div class="contents">
<div class="textblock"><code>#include <<a class="el" href="error__codes_8h_source.html">tvm/runtime/crt/error_codes.h</a>></code><br />
</div><div class="textblock"><div class="dynheader">
-Include dependency graph for graph_executor_module.h:</div>
+Include dependency graph for aot_executor_module.h:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="graph__executor__module_8h__incl.svg" width="170" height="156"><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="aot__executor__module_8h__incl.svg" width="211" height="142"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
</div>
-<p><a href="graph__executor__module_8h_source.html">Go to the source code of this file.</a></p>
+<p><a href="aot__executor__module_8h_source.html">Go to the source code of this file.</a></p>
<table class="memberdecls">
<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="func-members"></a>
Functions</h2></td></tr>
-<tr class="memitem:a174fff47eabdfcc1ecad5e7616020a35"><td class="memItemLeft" align="right" valign="top"><a class="el" href="error__codes_8h.html#a77b4da0131882f0c9b887a47dd34467a">tvm_crt_error_t</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="graph__executor__module_8h.html#a174fff47eabdfcc1ecad5e7616020a35">TVMGraphExecutorModule_Register</a> ()</td></tr>
-<tr class="memdesc:a174fff47eabdfcc1ecad5e7616020a35"><td class="mdescLeft"> </td><td class="mdescRight">Register the "tvm.graph_executor.create" constructor PackedFunc. <a href="#a174fff47eabdfcc1ecad5e7616020a35">More...</a><br /></td></tr>
-<tr class="separator:a174fff47eabdfcc1ecad5e7616020a35"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a7367944cb3bd7078b1e8a1630d0343d3"><td class="memItemLeft" align="right" valign="top"><a class="el" href="error__codes_8h.html#a77b4da0131882f0c9b887a47dd34467a">tvm_crt_error_t</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="aot__executor__module_8h.html#a7367944cb3bd7078b1e8a1630d0343d3">TVMAotExecutorModule_Register</a> ()</td></tr>
+<tr class="memdesc:a7367944cb3bd7078b1e8a1630d0343d3"><td class="mdescLeft"> </td><td class="mdescRight">Register the "tvm.aot_executor.create" constructor PackedFunc. <a href="#a7367944cb3bd7078b1e8a1630d0343d3">More...</a><br /></td></tr>
+<tr class="separator:a7367944cb3bd7078b1e8a1630d0343d3"><td class="memSeparator" colspan="2"> </td></tr>
</table>
<h2 class="groupheader">Function Documentation</h2>
-<a id="a174fff47eabdfcc1ecad5e7616020a35"></a>
-<h2 class="memtitle"><span class="permalink"><a href="#a174fff47eabdfcc1ecad5e7616020a35">◆ </a></span>TVMGraphExecutorModule_Register()</h2>
+<a id="a7367944cb3bd7078b1e8a1630d0343d3"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a7367944cb3bd7078b1e8a1630d0343d3">◆ </a></span>TVMAotExecutorModule_Register()</h2>
<div class="memitem">
<div class="memproto">
<table class="memname">
<tr>
- <td class="memname"><a class="el" href="error__codes_8h.html#a77b4da0131882f0c9b887a47dd34467a">tvm_crt_error_t</a> TVMGraphExecutorModule_Register </td>
+ <td class="memname"><a class="el" href="error__codes_8h.html#a77b4da0131882f0c9b887a47dd34467a">tvm_crt_error_t</a> TVMAotExecutorModule_Register </td>
<td>(</td>
<td class="paramname"></td><td>)</td>
<td></td>
@@ -100,7 +100,7 @@ Functions</h2></td></tr>
</table>
</div><div class="memdoc">
-<p>Register the "tvm.graph_executor.create" constructor PackedFunc. </p>
+<p>Register the "tvm.aot_executor.create" constructor PackedFunc. </p>
</div>
</div>
diff --git a/docs/reference/api/doxygen/aot__executor__module_8h__incl.svg b/docs/reference/api/doxygen/aot__executor__module_8h__incl.svg
new file mode 100644
index 000000000..1d4df8685
--- /dev/null
+++ b/docs/reference/api/doxygen/aot__executor__module_8h__incl.svg
@@ -0,0 +1,36 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.40.1 (20161225.0304)
+ -->
+<!-- Title: include/tvm/runtime/crt/aot_executor_module.h Pages: 1 -->
+<svg width="158pt" height="106pt"
+ viewBox="0.00 0.00 158.00 106.00" 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(4 102)">
+<title>include/tvm/runtime/crt/aot_executor_module.h</title>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-102 154,-102 154,4 -4,4"/>
+<!-- Node0 -->
+<g id="node1" class="node">
+<title>Node0</title>
+<polygon fill="#bfbfbf" stroke="#000000" points="0,-67.5 0,-97.5 150,-97.5 150,-67.5 0,-67.5"/>
+<text text-anchor="start" x="8" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="75" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/aot_executor_module.h</text>
+</g>
+<!-- Node1 -->
+<g id="node2" class="node">
+<title>Node1</title>
+<g id="a_node2"><a xlink:href="error__codes_8h.html" target="_top" xlink:title="Defines integral error codes returned by the CRT. ">
+<polygon fill="#ffffff" stroke="#000000" points="15.5,-.5 15.5,-30.5 134.5,-30.5 134.5,-.5 15.5,-.5"/>
+<text text-anchor="start" x="23.5" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm/runtime/crt/error</text>
+<text text-anchor="middle" x="75" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_codes.h</text>
+</a>
+</g>
+</g>
+<!-- Node0->Node1 -->
+<g id="edge1" class="edge">
+<title>Node0->Node1</title>
+<path fill="none" stroke="#191970" d="M75,-67.2967C75,-59.5013 75,-49.7991 75,-40.9064"/>
+<polygon fill="#191970" stroke="#191970" points="78.5001,-40.6431 75,-30.6432 71.5001,-40.6432 78.5001,-40.6431"/>
+</g>
+</g>
+</svg>
diff --git a/docs/reference/api/doxygen/aot__executor__module_8h_source.html b/docs/reference/api/doxygen/aot__executor__module_8h_source.html
new file mode 100644
index 000000000..c8315507e
--- /dev/null
+++ b/docs/reference/api/doxygen/aot__executor__module_8h_source.html
@@ -0,0 +1,80 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml">
+<head>
+<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
+<meta http-equiv="X-UA-Compatible" content="IE=9"/>
+<meta name="generator" content="Doxygen 1.8.13"/>
+<meta name="viewport" content="width=device-width, initial-scale=1"/>
+<title>tvm: include/tvm/runtime/crt/aot_executor_module.h Source File</title>
+<link href="tabs.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="jquery.js"></script>
+<script type="text/javascript" src="dynsections.js"></script>
+<link href="search/search.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="search/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+</head>
+<body>
+<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
+<div id="titlearea">
+<table cellspacing="0" cellpadding="0">
+ <tbody>
+ <tr style="height: 56px;">
+ <td id="projectalign" style="padding-left: 0.5em;">
+ <div id="projectname">tvm
+ </div>
+ </td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.13 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+<script type="text/javascript" src="menudata.js"></script>
+<script type="text/javascript" src="menu.js"></script>
+<script type="text/javascript">
+$(function() {
+ initMenu('',true,false,'search.php','Search');
+ $(document).ready(function() { init_search(); });
+});
+</script>
+<div id="main-nav"></div>
+<!-- window showing the filter options -->
+<div id="MSearchSelectWindow"
+ onmouseover="return searchBox.OnSearchSelectShow()"
+ onmouseout="return searchBox.OnSearchSelectHide()"
+ onkeydown="return searchBox.OnSearchSelectKey(event)">
+</div>
+
+<!-- iframe showing the search results (closed by default) -->
+<div id="MSearchResultsWindow">
+<iframe src="javascript:void(0)" frameborder="0"
+ name="MSearchResults" id="MSearchResults">
+</iframe>
+</div>
+
+<div id="nav-path" class="navpath">
+ <ul>
+<li class="navelem"><a class="el" href="dir_d44c64559bbebec7f509842c48db8b23.html">include</a></li><li class="navelem"><a class="el" href="dir_b4c7d8e826c599ba55146c099a14beb5.html">tvm</a></li><li class="navelem"><a class="el" href="dir_404558507ed35459f0d7a6d81d8c508d.html">runtime</a></li><li class="navelem"><a class="el" href="dir_a2900df4deca8dd2bcded616f0fe650a.html">crt</a></li> </ul>
+</div>
+</div><!-- top -->
+<div class="header">
+ <div class="headertitle">
+<div class="title">aot_executor_module.h</div> </div>
+</div><!--header-->
+<div class="contents">
+<a href="aot__executor__module_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">/*</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment"> * Licensed to the Apache Software Foundation (ASF) under one</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> [...]
+<div class="ttc" id="error__codes_8h_html"><div class="ttname"><a href="error__codes_8h.html">error_codes.h</a></div><div class="ttdoc">Defines integral error codes returned by the CRT. </div></div>
+<div class="ttc" id="error__codes_8h_html_a77b4da0131882f0c9b887a47dd34467a"><div class="ttname"><a href="error__codes_8h.html#a77b4da0131882f0c9b887a47dd34467a">tvm_crt_error_t</a></div><div class="ttdeci">tvm_crt_error_t</div><div class="ttdef"><b>Definition:</b> error_codes.h:50</div></div>
+</div><!-- fragment --></div><!-- contents -->
+<!-- start footer part -->
+<hr class="footer"/><address class="footer"><small>
+Generated by  <a href="http://www.doxygen.org/index.html">
+<img class="footer" src="doxygen.png" alt="doxygen"/>
+</a> 1.8.13
+</small></address>
+</body>
+</html>
diff --git a/docs/reference/api/doxygen/array_8h__dep__incl.svg b/docs/reference/api/doxygen/array_8h__dep__incl.svg
index ca6f80c8a..5a8496f85 100644
--- a/docs/reference/api/doxygen/array_8h__dep__incl.svg
+++ b/docs/reference/api/doxygen/array_8h__dep__incl.svg
@@ -91,39 +91,39 @@
<path fill="none" stroke="#191970" d="M1493.7806,-1016.6683C1386.4684,-1008.5613 1183.3117,-991.0252 1114,-970 952.2501,-920.9345 777.9995,-807.8823 728.0777,-774.1632"/>
<polygon fill="#191970" stroke="#191970" points="1493.6252,-1020.1663 1503.8587,-1017.4232 1494.1481,-1013.1859 1493.6252,-1020.1663"/>
</g>
-<!-- Node176 -->
+<!-- Node177 -->
<g id="node39" class="node">
-<title>Node176</title>
+<title>Node177</title>
<g id="a_node39"><a xlink:href="buffer_8h.html" target="_top" xlink:title="Symbolic n-dimensional array, to represent a memory buffer. ">
<polygon fill="#ffffff" stroke="#ff0000" points="2065,-565 2065,-584 2193,-584 2193,-565 2065,-565"/>
<text text-anchor="middle" x="2129" y="-572" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/buffer.h</text>
</a>
</g>
</g>
-<!-- Node19->Node176 -->
+<!-- Node19->Node177 -->
<g id="edge141" class="edge">
-<title>Node19->Node176</title>
+<title>Node19->Node177</title>
<path fill="none" stroke="#191970" d="M1630.4994,-1018.4843C1840.3197,-1008.3444 2460,-971.3034 2460,-887.5 2460,-887.5 2460,-887.5 2460,-703 2460,-644.4384 2277.6712,-602.021 2183.6579,-584.0419"/>
<polygon fill="#191970" stroke="#191970" points="1630.0499,-1015.0017 1620.2284,-1018.9749 1630.384,-1021.9937 1630.0499,-1015.0017"/>
</g>
-<!-- Node177 -->
+<!-- Node178 -->
<g id="node40" class="node">
-<title>Node177</title>
+<title>Node178</title>
<g id="a_node40"><a xlink:href="tir_2expr_8h.html" target="_top" xlink:title="TIR expressions. ">
<polygon fill="#ffffff" stroke="#ff0000" points="2023.5,-447.5 2023.5,-466.5 2144.5,-466.5 2144.5,-447.5 2023.5,-447.5"/>
<text text-anchor="middle" x="2084" y="-454.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/expr.h</text>
</a>
</g>
</g>
-<!-- Node19->Node177 -->
+<!-- Node19->Node178 -->
<g id="edge142" class="edge">
-<title>Node19->Node177</title>
+<title>Node19->Node178</title>
<path fill="none" stroke="#191970" d="M1630.1904,-1019.5326C1788.7778,-1014.5127 2178.5565,-999.4928 2306,-970 2396.4861,-949.0599 2498,-980.3775 2498,-887.5 2498,-887.5 2498,-887.5 2498,-574.5 2498,-501.9343 2259.635,-471.8408 2144.6425,-461.5451"/>
<polygon fill="#191970" stroke="#191970" points="1629.9697,-1016.0377 1620.0841,-1019.8488 1630.1887,-1023.0343 1629.9697,-1016.0377"/>
</g>
-<!-- Node182 -->
+<!-- Node183 -->
<g id="node41" class="node">
-<title>Node182</title>
+<title>Node183</title>
<g id="a_node41"><a xlink:href="index__map_8h.html" target="_top" xlink:title="Defines a remapping of buffer indices. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1552,-626.5 1552,-656.5 1670,-656.5 1670,-626.5 1552,-626.5"/>
<text text-anchor="start" x="1560" y="-644.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/index</text>
@@ -131,15 +131,15 @@
</a>
</g>
</g>
-<!-- Node19->Node182 -->
+<!-- Node19->Node183 -->
<g id="edge143" class="edge">
-<title>Node19->Node182</title>
+<title>Node19->Node183</title>
<path fill="none" stroke="#191970" d="M1562,-996.1092C1562,-969.1594 1562,-925.3265 1562,-887.5 1562,-887.5 1562,-887.5 1562,-764.5 1562,-723.1247 1586.2249,-679.1241 1600.6084,-656.6453"/>
<polygon fill="#191970" stroke="#191970" points="1558.5001,-996.227 1562,-1006.2271 1565.5001,-996.2271 1558.5001,-996.227"/>
</g>
-<!-- Node186 -->
+<!-- Node187 -->
<g id="node42" class="node">
-<title>Node186</title>
+<title>Node187</title>
<g id="a_node42"><a xlink:href="structural__equal_8h.html" target="_top" xlink:title="Structural equality comparison. ">
<polygon fill="#ffffff" stroke="#000000" points="1339.5,-939.5 1339.5,-969.5 1490.5,-969.5 1490.5,-939.5 1339.5,-939.5"/>
<text text-anchor="start" x="1347.5" y="-957.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/structural</text>
@@ -147,15 +147,15 @@
</a>
</g>
</g>
-<!-- Node19->Node186 -->
+<!-- Node19->Node187 -->
<g id="edge115" class="edge">
-<title>Node19->Node186</title>
+<title>Node19->Node187</title>
<path fill="none" stroke="#191970" d="M1519.9036,-1002.3132C1497.1153,-991.9267 1469.3667,-979.2794 1448.1336,-969.6017"/>
<polygon fill="#191970" stroke="#191970" points="1518.4593,-1005.5013 1529.0103,-1006.4639 1521.3625,-999.1317 1518.4593,-1005.5013"/>
</g>
-<!-- Node195 -->
+<!-- Node196 -->
<g id="node46" class="node">
-<title>Node195</title>
+<title>Node196</title>
<g id="a_node46"><a xlink:href="papi_8h.html" target="_top" xlink:title="include/tvm/runtime\l/contrib/papi.h">
<polygon fill="#ffffff" stroke="#000000" points="2526,-939.5 2526,-969.5 2642,-969.5 2642,-939.5 2526,-939.5"/>
<text text-anchor="start" x="2534" y="-957.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
@@ -163,15 +163,15 @@
</a>
</g>
</g>
-<!-- Node19->Node195 -->
+<!-- Node19->Node196 -->
<g id="edge127" class="edge">
-<title>Node19->Node195</title>
+<title>Node19->Node196</title>
<path fill="none" stroke="#191970" d="M1630.2512,-1019.2026C1818.3071,-1012.6489 2341.2992,-992.8553 2512,-970 2516.5624,-969.3891 2521.2654,-968.6436 2525.9793,-967.8132"/>
<polygon fill="#191970" stroke="#191970" points="1630.0882,-1015.7061 1620.2156,-1019.5509 1630.331,-1022.7019 1630.0882,-1015.7061"/>
</g>
-<!-- Node196 -->
+<!-- Node197 -->
<g id="node47" class="node">
-<title>Node196</title>
+<title>Node197</title>
<g id="a_node47"><a xlink:href="packed__func_8h.html" target="_top" xlink:title="Type-erased function used across TVM API. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1590,-939.5 1590,-969.5 1706,-969.5 1706,-939.5 1590,-939.5"/>
<text text-anchor="start" x="1598" y="-957.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
@@ -179,9 +179,9 @@
</a>
</g>
</g>
-<!-- Node19->Node196 -->
+<!-- Node19->Node197 -->
<g id="edge128" class="edge">
-<title>Node19->Node196</title>
+<title>Node19->Node197</title>
<path fill="none" stroke="#191970" d="M1589.5145,-1000.0643C1602.2304,-990.1577 1617.0316,-978.6265 1628.5625,-969.6432"/>
<polygon fill="#191970" stroke="#191970" points="1587.2523,-997.3899 1581.5147,-1006.2967 1591.5544,-1002.9119 1587.2523,-997.3899"/>
</g>
@@ -1020,21 +1020,21 @@
<path fill="none" stroke="#191970" d="M934.1877,-700.9628C794.9284,-696.4219 469.8265,-683.7004 198,-657 181.1528,-655.3452 162.9366,-653.025 146.2277,-650.6827"/>
<polygon fill="#191970" stroke="#191970" points="934.1821,-704.4643 944.2897,-701.2884 934.4077,-697.4679 934.1821,-704.4643"/>
</g>
-<!-- Node156->Node176 -->
+<!-- Node156->Node177 -->
<g id="edge95" class="edge">
-<title>Node156->Node176</title>
+<title>Node156->Node177</title>
<path fill="none" stroke="#191970" d="M1071.9482,-694.887C1143.7437,-686.2524 1259.4164,-671.817 1359,-657 1441.0273,-644.7952 1460.7406,-636.5273 1543,-626 1731.9283,-601.8216 1956.5929,-585.539 2064.654,-578.4801"/>
<polygon fill="#191970" stroke="#191970" points="1071.1931,-691.4524 1061.681,-696.1182 1072.0266,-698.4026 1071.1931,-691.4524"/>
</g>
-<!-- Node156->Node177 -->
+<!-- Node156->Node178 -->
<g id="edge104" class="edge">
-<title>Node156->Node177</title>
+<title>Node156->Node178</title>
<path fill="none" stroke="#191970" d="M1071.7743,-696.0649C1129.0379,-689.2569 1212.3337,-676.9432 1283,-657 1318.1083,-647.0919 1324.2387,-637.0643 1359,-626 1605.0012,-547.6995 1908.5053,-488.6989 2030.0043,-466.5512"/>
<polygon fill="#191970" stroke="#191970" points="1071.2778,-692.5989 1061.7497,-697.232 1072.0874,-699.552 1071.2778,-692.5989"/>
</g>
-<!-- Node156->Node182 -->
+<!-- Node156->Node183 -->
<g id="edge105" class="edge">
-<title>Node156->Node182</title>
+<title>Node156->Node183</title>
<path fill="none" stroke="#191970" d="M1072.0492,-699.7925C1174.2785,-694.4275 1371.734,-681.703 1538,-657 1542.4854,-656.3336 1547.1089,-655.5597 1551.7488,-654.7201"/>
<polygon fill="#191970" stroke="#191970" points="1071.6645,-696.3077 1061.8586,-700.3194 1072.026,-703.2983 1071.6645,-696.3077"/>
</g>
@@ -1092,123 +1092,123 @@
<path fill="none" stroke="#191970" d="M92.683,-616.1079C97.9578,-590.2919 105,-549.0711 105,-513 105,-513 105,-513 105,-457 105,-409.7256 134.3121,-358.7094 147.5941,-338.0193"/>
<polygon fill="#191970" stroke="#191970" points="89.2001,-615.6612 90.5562,-626.1688 96.0488,-617.1089 89.2001,-615.6612"/>
</g>
-<!-- Node176->Node150 -->
+<!-- Node177->Node150 -->
<g id="edge103" class="edge">
-<title>Node176->Node150</title>
+<title>Node177->Node150</title>
<path fill="none" stroke="#191970" d="M2137.0526,-555.5999C2146.638,-530.7559 2160.5723,-485.7235 2154,-447 2151.4326,-431.8727 2144.9064,-415.2677 2140.4599,-405.1129"/>
<polygon fill="#191970" stroke="#191970" points="2133.7824,-554.3509 2133.2834,-564.934 2140.2732,-556.9721 2133.7824,-554.3509"/>
</g>
-<!-- Node176->Node177 -->
+<!-- Node177->Node178 -->
<g id="edge96" class="edge">
-<title>Node176->Node177</title>
+<title>Node177->Node178</title>
<path fill="none" stroke="#191970" d="M2121.6279,-555.2507C2111.9047,-529.8622 2095.1597,-486.1393 2087.7334,-466.7484"/>
<polygon fill="#191970" stroke="#191970" points="2118.4724,-556.7977 2125.3175,-564.8845 2125.0094,-554.2941 2118.4724,-556.7977"/>
</g>
-<!-- Node177->Node54 -->
+<!-- Node178->Node54 -->
<g id="edge98" class="edge">
-<title>Node177->Node54</title>
+<title>Node178->Node54</title>
<path fill="none" stroke="#191970" d="M2154.4075,-445.5141C2220.1789,-431.5712 2317.6231,-402.279 2382,-344 2424.1037,-305.8844 2447.1878,-239.6829 2455.8996,-209.8523"/>
<polygon fill="#191970" stroke="#191970" points="2153.6286,-442.1009 2144.5376,-447.5419 2155.0374,-448.9577 2153.6286,-442.1009"/>
</g>
-<!-- Node177->Node92 -->
+<!-- Node178->Node92 -->
<g id="edge99" class="edge">
-<title>Node177->Node92</title>
+<title>Node178->Node92</title>
<path fill="none" stroke="#191970" d="M2068.3978,-439.7166C2043.7003,-412.3577 1996.3979,-359.9581 1976.82,-338.2705"/>
<polygon fill="#191970" stroke="#191970" points="2065.9823,-442.264 2075.2811,-447.3416 2071.1783,-437.5734 2065.9823,-442.264"/>
</g>
-<!-- Node177->Node108 -->
+<!-- Node178->Node108 -->
<g id="edge102" class="edge">
-<title>Node177->Node108</title>
+<title>Node178->Node108</title>
<path fill="none" stroke="#191970" d="M2069.3765,-438.7107C2058.4887,-422.7217 2047.082,-399.1347 2057,-380 2067.2779,-360.1709 2089.0909,-346.3546 2106.2099,-338.0608"/>
<polygon fill="#191970" stroke="#191970" points="2066.8294,-441.1565 2075.562,-447.1558 2072.4767,-437.0202 2066.8294,-441.1565"/>
</g>
-<!-- Node177->Node115 -->
+<!-- Node178->Node115 -->
<g id="edge101" class="edge">
-<title>Node177->Node115</title>
+<title>Node178->Node115</title>
<path fill="none" stroke="#191970" d="M2013.2181,-452.9313C1910.4634,-446.4062 1725.3384,-432.1489 1662,-411 1581.098,-383.9865 1575.1871,-347.0829 1497,-313 1453.7426,-294.1435 1401.5565,-279.791 1365.4891,-271.0874"/>
<polygon fill="#191970" stroke="#191970" points="2013.0022,-456.4245 2023.2017,-453.5574 2013.4404,-449.4383 2013.0022,-456.4245"/>
</g>
-<!-- Node177->Node137 -->
+<!-- Node178->Node137 -->
<g id="edge97" class="edge">
-<title>Node177->Node137</title>
+<title>Node178->Node137</title>
<path fill="none" stroke="#191970" d="M2013.3211,-452.3364C1780.363,-436.9585 1015.5058,-386.4149 383,-344 328.9283,-340.374 267.6779,-336.2261 222.1366,-333.1342"/>
<polygon fill="#191970" stroke="#191970" points="2013.147,-455.8325 2023.3559,-452.9988 2013.6082,-448.8477 2013.147,-455.8325"/>
</g>
-<!-- Node177->Node150 -->
+<!-- Node178->Node150 -->
<g id="edge100" class="edge">
-<title>Node177->Node150</title>
+<title>Node178->Node150</title>
<path fill="none" stroke="#191970" d="M2098.7338,-439.5744C2108.1384,-428.4517 2119.9811,-414.4454 2127.7604,-405.2449"/>
<polygon fill="#191970" stroke="#191970" points="2095.909,-437.4945 2092.125,-447.3906 2101.2544,-442.0142 2095.909,-437.4945"/>
</g>
-<!-- Node186->Node161 -->
+<!-- Node187->Node161 -->
<g id="edge116" class="edge">
-<title>Node186->Node161</title>
+<title>Node187->Node161</title>
<path fill="none" stroke="#191970" d="M1329.1524,-948.9452C1170.6619,-936.1456 824.0227,-896.1642 558,-780 510.9832,-759.4692 494.5438,-754.2211 464,-713 449.8124,-693.8528 443.6674,-665.7921 441.3003,-651.1694"/>
<polygon fill="#191970" stroke="#191970" points="1329.072,-952.4497 1339.317,-949.7495 1329.6243,-945.4715 1329.072,-952.4497"/>
</g>
-<!-- Node187 -->
+<!-- Node188 -->
<g id="node43" class="node">
-<title>Node187</title>
+<title>Node188</title>
<g id="a_node43"><a xlink:href="node_8h.html" target="_top" xlink:title="Definitions and helper macros for IR/AST nodes. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1386.5,-816.5 1386.5,-835.5 1523.5,-835.5 1523.5,-816.5 1386.5,-816.5"/>
<text text-anchor="middle" x="1455" y="-823.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/node.h</text>
</a>
</g>
</g>
-<!-- Node186->Node187 -->
+<!-- Node187->Node188 -->
<g id="edge117" class="edge">
-<title>Node186->Node187</title>
+<title>Node187->Node188</title>
<path fill="none" stroke="#191970" d="M1387.7474,-932.0412C1370.9616,-915.3804 1354.7913,-892.1579 1367,-872 1377.9848,-853.863 1398.7287,-842.5354 1417.318,-835.6607"/>
<polygon fill="#191970" stroke="#191970" points="1385.7085,-934.9265 1395.4033,-939.1998 1390.4894,-929.8134 1385.7085,-934.9265"/>
</g>
-<!-- Node190 -->
+<!-- Node191 -->
<g id="node45" class="node">
-<title>Node190</title>
+<title>Node191</title>
<g id="a_node45"><a xlink:href="reflection_8h.html" target="_top" xlink:title="Reflection and serialization of compiler IR/AST nodes. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1375.5,-878 1375.5,-897 1534.5,-897 1534.5,-878 1375.5,-878"/>
<text text-anchor="middle" x="1455" y="-885" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/reflection.h</text>
</a>
</g>
</g>
-<!-- Node186->Node190 -->
+<!-- Node187->Node191 -->
<g id="edge125" class="edge">
-<title>Node186->Node190</title>
+<title>Node187->Node191</title>
<path fill="none" stroke="#191970" d="M1429.3283,-930.5001C1436.2265,-918.9456 1444.1061,-905.7473 1449.2955,-897.055"/>
<polygon fill="#191970" stroke="#191970" points="1426.1976,-928.9163 1424.0766,-939.2967 1432.2079,-932.5046 1426.1976,-928.9163"/>
</g>
-<!-- Node187->Node20 -->
+<!-- Node188->Node20 -->
<g id="edge119" class="edge">
-<title>Node187->Node20</title>
+<title>Node188->Node20</title>
<path fill="none" stroke="#191970" d="M1376.2162,-815.6598C1319.1832,-807.5764 1240.4202,-795.1797 1172,-780 1065.1369,-756.2914 1039.7583,-744.7417 935,-713 865.5319,-691.9512 784.4943,-665.1538 742.5007,-651.0979"/>
<polygon fill="#191970" stroke="#191970" points="1375.8576,-819.1437 1386.2473,-817.0688 1376.8314,-812.2117 1375.8576,-819.1437"/>
</g>
-<!-- Node187->Node29 -->
+<!-- Node188->Node29 -->
<g id="edge123" class="edge">
-<title>Node187->Node29</title>
+<title>Node188->Node29</title>
<path fill="none" stroke="#191970" d="M1456.5788,-806.0747C1461.9725,-747.317 1483.4943,-571.098 1555,-447 1580.3279,-403.0436 1626.2007,-364.0964 1652.9893,-343.6557"/>
<polygon fill="#191970" stroke="#191970" points="1453.077,-805.9462 1455.7051,-816.2099 1460.0512,-806.5475 1453.077,-805.9462"/>
</g>
-<!-- Node187->Node30 -->
+<!-- Node188->Node30 -->
<g id="edge122" class="edge">
-<title>Node187->Node30</title>
+<title>Node188->Node30</title>
<path fill="none" stroke="#191970" d="M1473.8761,-809.8772C1492.6344,-794.0134 1522.3709,-769.2894 1549,-749 1605.3022,-706.102 1644.3455,-718.719 1679,-657 1724.7826,-575.462 1726.6475,-533.8565 1692,-447 1687.6926,-436.2018 1572.5659,-318.8456 1530.8846,-276.5686"/>
<polygon fill="#191970" stroke="#191970" points="1471.4391,-807.3549 1466.0805,-816.4947 1475.9692,-812.6914 1471.4391,-807.3549"/>
</g>
-<!-- Node187->Node154 -->
+<!-- Node188->Node154 -->
<g id="edge121" class="edge">
-<title>Node187->Node154</title>
+<title>Node188->Node154</title>
<path fill="none" stroke="#191970" d="M1376.2645,-819.619C1267.8808,-810.8155 1066.6257,-794.3975 895,-780 854.3004,-776.5858 808.4606,-772.6594 773.0958,-769.6128"/>
<polygon fill="#191970" stroke="#191970" points="1376.2087,-823.1259 1386.4592,-820.4469 1376.7753,-816.1489 1376.2087,-823.1259"/>
</g>
-<!-- Node187->Node156 -->
+<!-- Node188->Node156 -->
<g id="edge120" class="edge">
-<title>Node187->Node156</title>
+<title>Node188->Node156</title>
<path fill="none" stroke="#191970" d="M1410.0022,-813.755C1319.9439,-789.248 1119.2348,-734.6303 1038.0692,-712.5432"/>
<polygon fill="#191970" stroke="#191970" points="1409.3403,-817.2021 1419.9084,-816.4507 1411.1784,-810.4477 1409.3403,-817.2021"/>
</g>
-<!-- Node187->Node177 -->
+<!-- Node188->Node178 -->
<g id="edge124" class="edge">
-<title>Node187->Node177</title>
+<title>Node188->Node178</title>
<path fill="none" stroke="#191970" d="M1484.5412,-812.1205C1543.4552,-784.4018 1671.5865,-723.9091 1691,-713 1845.3988,-626.2386 2021.1219,-502.1138 2070.484,-466.7416"/>
<polygon fill="#191970" stroke="#191970" points="1482.9358,-809.0077 1475.3762,-816.4308 1485.9149,-815.3421 1482.9358,-809.0077"/>
</g>
@@ -1222,57 +1222,57 @@
</a>
</g>
</g>
-<!-- Node187->Node34 -->
+<!-- Node188->Node34 -->
<g id="edge118" class="edge">
-<title>Node187->Node34</title>
+<title>Node188->Node34</title>
<path fill="none" stroke="#191970" d="M1503.3828,-814.05C1543.2362,-804.2067 1599.9763,-790.1926 1642.9508,-779.5784"/>
<polygon fill="#191970" stroke="#191970" points="1502.5021,-810.6623 1493.633,-816.4581 1504.1806,-817.4581 1502.5021,-810.6623"/>
</g>
-<!-- Node190->Node187 -->
+<!-- Node191->Node188 -->
<g id="edge126" class="edge">
-<title>Node190->Node187</title>
+<title>Node191->Node188</title>
<path fill="none" stroke="#191970" d="M1455,-867.7462C1455,-857.0519 1455,-844.314 1455,-835.7449"/>
<polygon fill="#191970" stroke="#191970" points="1451.5001,-867.8906 1455,-877.8906 1458.5001,-867.8907 1451.5001,-867.8906"/>
</g>
-<!-- Node196->Node22 -->
+<!-- Node197->Node22 -->
<g id="edge130" class="edge">
-<title>Node196->Node22</title>
+<title>Node197->Node22</title>
<path fill="none" stroke="#191970" d="M1716.0698,-949.7441C1898.3879,-936.1006 2384,-893.4313 2384,-826 2384,-826 2384,-826 2384,-703 2384,-562.221 2460,-536.279 2460,-395.5 2460,-395.5 2460,-395.5 2460,-328.5 2460,-219.9441 2072.4148,-199.3294 1928.2059,-195.4163"/>
<polygon fill="#191970" stroke="#191970" points="1715.7735,-946.2564 1706.0599,-950.4868 1716.2915,-953.2372 1715.7735,-946.2564"/>
</g>
-<!-- Node196->Node26 -->
+<!-- Node197->Node26 -->
<g id="edge133" class="edge">
-<title>Node196->Node26</title>
+<title>Node197->Node26</title>
<path fill="none" stroke="#191970" d="M1645.5063,-928.9037C1645.0385,-911.6768 1646.8701,-889.0775 1657,-872 1697.91,-803.0319 1816.5393,-824.3114 1789,-749 1779.7512,-723.7075 1511.347,-416.7017 1447.5226,-343.9071"/>
<polygon fill="#191970" stroke="#191970" points="1642.0362,-929.5444 1646.0817,-939.3364 1649.0256,-929.1588 1642.0362,-929.5444"/>
</g>
-<!-- Node196->Node54 -->
+<!-- Node197->Node54 -->
<g id="edge140" class="edge">
-<title>Node196->Node54</title>
+<title>Node197->Node54</title>
<path fill="none" stroke="#191970" d="M1716.2069,-951.3366C1850.5495,-944.6558 2144.8747,-927.668 2243,-903 2326.9906,-881.8853 2422,-912.604 2422,-826 2422,-826 2422,-826 2422,-703 2422,-604.5217 2536,-611.4783 2536,-513 2536,-513 2536,-513 2536,-328.5 2536,-280.0913 2498.0081,-233.1499 2475.7472,-209.8"/>
<polygon fill="#191970" stroke="#191970" points="1715.8548,-947.8496 1706.0393,-951.8377 1716.1995,-954.8411 1715.8548,-947.8496"/>
</g>
-<!-- Node196->Node161 -->
+<!-- Node197->Node161 -->
<g id="edge131" class="edge">
-<title>Node196->Node161</title>
+<title>Node197->Node161</title>
<path fill="none" stroke="#191970" d="M1579.6034,-946.246C1382.8699,-921.9307 818.5926,-848.0059 646,-780 563.0028,-747.297 479.5399,-676.9263 450.63,-651.1856"/>
<polygon fill="#191970" stroke="#191970" points="1579.4532,-949.7539 1589.8062,-947.5032 1580.3094,-942.8064 1579.4532,-949.7539"/>
</g>
-<!-- Node196->Node34 -->
+<!-- Node197->Node34 -->
<g id="edge129" class="edge">
-<title>Node196->Node34</title>
+<title>Node197->Node34</title>
<path fill="none" stroke="#191970" d="M1633.2649,-930.5039C1624.9991,-913.9648 1617.5302,-891.5308 1624,-872 1636.5096,-834.2365 1668.0017,-798.9581 1687.5354,-779.6942"/>
<polygon fill="#191970" stroke="#191970" points="1630.2902,-932.3614 1638.1243,-939.4943 1636.4482,-929.0328 1630.2902,-932.3614"/>
</g>
-<!-- Node196->Node190 -->
+<!-- Node197->Node191 -->
<g id="edge132" class="edge">
-<title>Node196->Node190</title>
+<title>Node197->Node191</title>
<path fill="none" stroke="#191970" d="M1595.1764,-936.1623C1558.6806,-923.4928 1511.6218,-907.1563 1482.3809,-897.0053"/>
<polygon fill="#191970" stroke="#191970" points="1594.0922,-939.4907 1604.687,-939.4639 1596.3879,-932.8779 1594.0922,-939.4907"/>
</g>
-<!-- Node189 -->
+<!-- Node190 -->
<g id="node48" class="node">
-<title>Node189</title>
+<title>Node190</title>
<g id="a_node48"><a xlink:href="source__map_8h.html" target="_top" xlink:title="A map from source names to source code. ">
<polygon fill="#ffffff" stroke="#000000" points="1171,-559.5 1171,-589.5 1279,-589.5 1279,-559.5 1171,-559.5"/>
<text text-anchor="start" x="1179" y="-577.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/parser</text>
@@ -1280,15 +1280,15 @@
</a>
</g>
</g>
-<!-- Node196->Node189 -->
+<!-- Node197->Node190 -->
<g id="edge134" class="edge">
-<title>Node196->Node189</title>
+<title>Node197->Node190</title>
<path fill="none" stroke="#191970" d="M1579.4339,-947.7583C1507.9704,-939.7186 1401.6508,-924.5995 1367,-903 1330.8666,-880.4763 1310,-868.5786 1310,-826 1310,-826 1310,-826 1310,-764.5 1310,-694.0337 1260.6817,-620.6775 1237.1784,-589.7196"/>
<polygon fill="#191970" stroke="#191970" points="1579.3884,-951.2745 1589.7121,-948.8931 1580.1566,-944.3168 1579.3884,-951.2745"/>
</g>
-<!-- Node197 -->
+<!-- Node198 -->
<g id="node49" class="node">
-<title>Node197</title>
+<title>Node198</title>
<g id="a_node49"><a xlink:href="device__api_8h.html" target="_top" xlink:title="Abstract device memory management API. ">
<polygon fill="#ffffff" stroke="#ff0000" points="1666,-872.5 1666,-902.5 1782,-902.5 1782,-872.5 1666,-872.5"/>
<text text-anchor="start" x="1674" y="-890.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
@@ -1296,15 +1296,15 @@
</a>
</g>
</g>
-<!-- Node196->Node197 -->
+<!-- Node197->Node198 -->
<g id="edge137" class="edge">
-<title>Node196->Node197</title>
+<title>Node197->Node198</title>
<path fill="none" stroke="#191970" d="M1672.9592,-932.4965C1684.0497,-922.7193 1696.8267,-911.4554 1706.8227,-902.6432"/>
<polygon fill="#191970" stroke="#191970" points="1670.4323,-930.0582 1665.2456,-939.2967 1675.0614,-935.3091 1670.4323,-930.0582"/>
</g>
-<!-- Node199 -->
+<!-- Node200 -->
<g id="node50" class="node">
-<title>Node199</title>
+<title>Node200</title>
<g id="a_node50"><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="#ff0000" points="1800,-872.5 1800,-902.5 1916,-902.5 1916,-872.5 1800,-872.5"/>
<text text-anchor="start" x="1808" y="-890.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
@@ -1312,27 +1312,27 @@
</a>
</g>
</g>
-<!-- Node196->Node199 -->
+<!-- Node197->Node200 -->
<g id="edge138" class="edge">
-<title>Node196->Node199</title>
+<title>Node197->Node200</title>
<path fill="none" stroke="#191970" d="M1699.0373,-936.2759C1731.1443,-925.7165 1772.4285,-912.5805 1804.9178,-902.6017"/>
<polygon fill="#191970" stroke="#191970" points="1697.7849,-933.0035 1689.3876,-939.4639 1699.9808,-939.6502 1697.7849,-933.0035"/>
</g>
-<!-- Node189->Node21 -->
+<!-- Node190->Node21 -->
<g id="edge136" class="edge">
-<title>Node189->Node21</title>
+<title>Node190->Node21</title>
<path fill="none" stroke="#191970" d="M1205.8201,-551.3713C1197.6399,-541.5069 1188.5336,-530.5258 1182.1657,-522.8469"/>
<polygon fill="#191970" stroke="#191970" points="1203.3157,-553.8343 1212.3933,-559.2977 1208.704,-549.366 1203.3157,-553.8343"/>
</g>
-<!-- Node189->Node23 -->
+<!-- Node190->Node23 -->
<g id="edge135" class="edge">
-<title>Node189->Node23</title>
+<title>Node190->Node23</title>
<path fill="none" stroke="#191970" d="M1176.4862,-555.7488C1110.8653,-530.3854 996.2454,-486.0833 945.9078,-466.6272"/>
<polygon fill="#191970" stroke="#191970" points="1175.6,-559.1586 1186.1894,-559.4992 1178.1237,-552.6293 1175.6,-559.1586"/>
</g>
-<!-- Node199->Node196 -->
+<!-- Node200->Node197 -->
<g id="edge139" class="edge">
-<title>Node199->Node196</title>
+<title>Node200->Node197</title>
<path fill="none" stroke="#191970" d="M1806.7514,-905.7936C1774.6042,-916.3642 1733.3193,-929.499 1700.8686,-939.4639"/>
<polygon fill="#191970" stroke="#191970" points="1808.0172,-909.0615 1816.4149,-902.6017 1805.8217,-902.4147 1808.0172,-909.0615"/>
</g>
diff --git a/docs/reference/api/doxygen/c__backend__api_8h__dep__incl.svg b/docs/reference/api/doxygen/c__backend__api_8h__dep__incl.svg
index a2037d4a9..73657217d 100644
--- a/docs/reference/api/doxygen/c__backend__api_8h__dep__incl.svg
+++ b/docs/reference/api/doxygen/c__backend__api_8h__dep__incl.svg
@@ -73,7 +73,7 @@
<!-- Node9 -->
<g id="node5" class="node">
<title>Node9</title>
-<g id="a_node5"><a xlink:href="graph__executor_8h.html" target="_top" xlink:title="Tiny graph executor that can run graph containing only tvm PackedFunc. ">
+<g id="a_node5"><a xlink:href="graph__executor_8h.html" target="_top" xlink:title="Tiny AoT executor. ">
<polygon fill="#ffffff" stroke="#000000" points="41.5,-.5 41.5,-30.5 160.5,-30.5 160.5,-.5 41.5,-.5"/>
<text text-anchor="start" x="49.5" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
<text text-anchor="middle" x="101" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/graph_executor.h</text>
diff --git a/docs/reference/api/doxygen/c__runtime__api_8h.html b/docs/reference/api/doxygen/c__runtime__api_8h.html
index 42dc6fffd..fe3506998 100644
--- a/docs/reference/api/doxygen/c__runtime__api_8h.html
+++ b/docs/reference/api/doxygen/c__runtime__api_8h.html
@@ -84,7 +84,7 @@ Include dependency graph for c_runtime_api.h:</div>
</div><div class="textblock"><div class="dynheader">
This graph shows which files directly or indirectly include this file:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="c__runtime__api_8h__dep__incl.svg" width="3160" height="1020"><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="c__runtime__api_8h__dep__incl.svg" width="2978" height="1020"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
</div>
@@ -455,7 +455,7 @@ Functions</h2></td></tr>
<tr><td class="paramname">type_codes</td><td>The type codes of the arguments </td></tr>
<tr><td class="paramname">num_args</td><td>Number of arguments. </td></tr>
<tr><td class="paramname">ret</td><td>The return value handle. </td></tr>
- <tr><td class="paramname">resource_handle</td><td>The handle additional resouce handle from fron-end. </td></tr>
+ <tr><td class="paramname">resource_handle</td><td>The handle additional resouce handle from front-end. </td></tr>
</table>
</dd>
</dl>
@@ -479,7 +479,7 @@ Functions</h2></td></tr>
<p>C callback to free the resource handle in C packed function. </p>
<dl class="params"><dt>Parameters</dt><dd>
<table class="params">
- <tr><td class="paramname">resource_handle</td><td>The handle additional resouce handle from fron-end. </td></tr>
+ <tr><td class="paramname">resource_handle</td><td>The handle additional resouce handle from front-end. </td></tr>
</table>
</dd>
</dl>
diff --git a/docs/reference/api/doxygen/c__runtime__api_8h__dep__incl.svg b/docs/reference/api/doxygen/c__runtime__api_8h__dep__incl.svg
index c2fde2664..b6556ceaf 100644
--- a/docs/reference/api/doxygen/c__runtime__api_8h__dep__incl.svg
+++ b/docs/reference/api/doxygen/c__runtime__api_8h__dep__incl.svg
@@ -4,1162 +4,1174 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: include/tvm/runtime/c_runtime_api.h Pages: 1 -->
-<svg width="2370pt" height="765pt"
- viewBox="0.00 0.00 2370.00 765.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<svg width="2233pt" height="765pt"
+ viewBox="0.00 0.00 2232.50 765.00" 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(4 761)">
<title>include/tvm/runtime/c_runtime_api.h</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-761 2366,-761 2366,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-761 2228.5,-761 2228.5,4 -4,4"/>
<!-- Node4 -->
<g id="node1" class="node">
<title>Node4</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="994,-726.5 994,-756.5 1110,-756.5 1110,-726.5 994,-726.5"/>
-<text text-anchor="start" x="1002" y="-744.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1052" y="-733.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/c_runtime_api.h</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="1033.5,-726.5 1033.5,-756.5 1149.5,-756.5 1149.5,-726.5 1033.5,-726.5"/>
+<text text-anchor="start" x="1041.5" y="-744.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1091.5" y="-733.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/c_runtime_api.h</text>
</g>
<!-- Node5 -->
<g id="node2" class="node">
<title>Node5</title>
<g id="a_node2"><a xlink:href="compute__dag_8h.html" target="_top" xlink:title="The auto-scheduler's computational graph and related program analyses. ">
-<polygon fill="#ffffff" stroke="#000000" points="0,-391.5 0,-421.5 152,-421.5 152,-391.5 0,-391.5"/>
-<text text-anchor="start" x="8" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="76" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/compute_dag.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="265.5,-391.5 265.5,-421.5 417.5,-421.5 417.5,-391.5 265.5,-391.5"/>
+<text text-anchor="start" x="273.5" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="341.5" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/compute_dag.h</text>
</a>
</g>
</g>
<!-- Node4->Node5 -->
<g id="edge1" class="edge">
<title>Node4->Node5</title>
-<path fill="none" stroke="#191970" d="M983.7701,-741.2102C781.915,-739.8699 197.0983,-732.336 123,-690 84.7371,-668.1385 68,-651.5678 68,-607.5 68,-607.5 68,-607.5 68,-540.5 68,-497.3852 72.1779,-446.6056 74.4953,-421.7553"/>
-<polygon fill="#191970" stroke="#191970" points="983.9251,-744.7111 993.9471,-741.2746 983.9695,-737.7113 983.9251,-744.7111"/>
+<path fill="none" stroke="#191970" d="M1023.1689,-740.6748C862.7438,-738.1115 472.1392,-727.9963 427.5,-690 398.7698,-665.5453 407.5,-645.2287 407.5,-607.5 407.5,-607.5 407.5,-607.5 407.5,-540.5 407.5,-493.3026 374.1415,-445.4861 354.8561,-421.7859"/>
+<polygon fill="#191970" stroke="#191970" points="1023.3439,-744.1779 1033.3968,-740.833 1023.4521,-737.1787 1023.3439,-744.1779"/>
</g>
<!-- Node13 -->
<g id="node7" class="node">
<title>Node13</title>
<g id="a_node7"><a xlink:href="node_8h.html" target="_top" xlink:title="Definitions and helper macros for IR/AST nodes. ">
-<polygon fill="#ffffff" stroke="#000000" points="1111.5,-397 1111.5,-416 1248.5,-416 1248.5,-397 1111.5,-397"/>
-<text text-anchor="middle" x="1180" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/node.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="760,-397 760,-416 897,-416 897,-397 760,-397"/>
+<text text-anchor="middle" x="828.5" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/node.h</text>
</a>
</g>
</g>
<!-- Node4->Node13 -->
<g id="edge6" class="edge">
<title>Node4->Node13</title>
-<path fill="none" stroke="#191970" d="M1093.1682,-721.5837C1145.8097,-693.3486 1233.7718,-636.0661 1264,-556 1285.2411,-499.7382 1220.661,-439.1829 1192.4712,-416.1395"/>
-<polygon fill="#191970" stroke="#191970" points="1091.3131,-718.6046 1084.0883,-726.354 1094.5688,-724.8014 1091.3131,-718.6046"/>
+<path fill="none" stroke="#191970" d="M1023.4485,-735.3437C937.1444,-726.8582 793.6401,-710.2906 744.5,-690 723.0555,-681.1453 723.7852,-668.231 702.5,-659 621.6449,-623.9348 563.8782,-690.7392 507.5,-623 419.1328,-516.8256 679.4832,-441.6021 785.6382,-416.0557"/>
+<polygon fill="#191970" stroke="#191970" points="1023.1296,-738.8291 1033.4214,-736.3133 1023.8071,-731.8619 1023.1296,-738.8291"/>
</g>
-<!-- Node162 -->
+<!-- Node163 -->
<g id="node15" class="node">
-<title>Node162</title>
+<title>Node163</title>
<g id="a_node15"><a xlink:href="tir_2expr_8h.html" target="_top" xlink:title="TIR expressions. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="973.5,-73 973.5,-92 1094.5,-92 1094.5,-73 973.5,-73"/>
-<text text-anchor="middle" x="1034" y="-80" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/expr.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="854,-73 854,-92 975,-92 975,-73 854,-73"/>
+<text text-anchor="middle" x="914.5" y="-80" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/expr.h</text>
</a>
</g>
</g>
-<!-- Node4->Node162 -->
-<g id="edge111" class="edge">
-<title>Node4->Node162</title>
-<path fill="none" stroke="#191970" d="M983.685,-734.2436C901.2221,-724.8989 767.7859,-707.7392 721,-690 706.6191,-684.5474 488,-555.8799 488,-540.5 488,-540.5 488,-540.5 488,-406.5 488,-246.5733 580.7362,-207.0585 723,-134 814.649,-86.9343 850.4815,-115.57 952,-98 962.707,-96.1469 974.1785,-94.0493 985.0475,-92.0094"/>
-<polygon fill="#191970" stroke="#191970" points="983.3992,-737.7334 993.7274,-735.3715 984.1806,-730.7771 983.3992,-737.7334"/>
+<!-- Node4->Node163 -->
+<g id="edge113" class="edge">
+<title>Node4->Node163</title>
+<path fill="none" stroke="#191970" d="M1159.6897,-740.0217C1370.3305,-735.1609 1998.5101,-718.3425 2031.5,-690 2098.6758,-632.2875 2085.9258,-476.5449 2030.5,-324 1993.1684,-221.2548 1923.2942,-228.5283 1817.5,-201 1507.2529,-120.2719 1122.0241,-93.0865 975.1865,-85.2699"/>
+<polygon fill="#191970" stroke="#191970" points="1159.5982,-736.5228 1149.681,-740.2509 1159.7585,-743.5209 1159.5982,-736.5228"/>
</g>
-<!-- Node174 -->
+<!-- Node175 -->
<g id="node22" class="node">
-<title>Node174</title>
+<title>Node175</title>
<g id="a_node22"><a xlink:href="reflection_8h.html" target="_top" xlink:title="Reflection and serialization of compiler IR/AST nodes. ">
-<polygon fill="#ffffff" stroke="#000000" points="806.5,-464 806.5,-483 965.5,-483 965.5,-464 806.5,-464"/>
-<text text-anchor="middle" x="886" y="-471" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/reflection.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="953,-464 953,-483 1112,-483 1112,-464 953,-464"/>
+<text text-anchor="middle" x="1032.5" y="-471" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/reflection.h</text>
</a>
</g>
</g>
-<!-- Node4->Node174 -->
+<!-- Node4->Node175 -->
<g id="edge38" class="edge">
-<title>Node4->Node174</title>
-<path fill="none" stroke="#191970" d="M983.8047,-734.3152C911.2011,-725.7843 801.7906,-710.1392 765,-690 729.4119,-670.519 722.8588,-658.3787 703,-623 681.2835,-584.3117 656.7777,-560.0334 684,-525 699.7386,-504.7454 759.9803,-491.1315 810.616,-483.0142"/>
-<polygon fill="#191970" stroke="#191970" points="983.4576,-737.7983 993.7937,-735.4715 984.2625,-730.8448 983.4576,-737.7983"/>
+<title>Node4->Node175</title>
+<path fill="none" stroke="#191970" d="M1023.2626,-729.6569C980.9515,-720.7667 932.2948,-707.1871 919.5,-690 891.7134,-652.6748 994.651,-520.1917 1024.5691,-483.1721"/>
+<polygon fill="#191970" stroke="#191970" points="1022.7967,-733.1339 1033.2946,-731.704 1024.1964,-726.2752 1022.7967,-733.1339"/>
</g>
-<!-- Node179 -->
+<!-- Node180 -->
<g id="node26" class="node">
-<title>Node179</title>
+<title>Node180</title>
<g id="a_node26"><a xlink:href="serialization_8h.html" target="_top" xlink:title="include/tvm/node/serialization.h">
-<polygon fill="#ffffff" stroke="#000000" points="1290.5,-598 1290.5,-617 1463.5,-617 1463.5,-598 1290.5,-598"/>
-<text text-anchor="middle" x="1377" y="-605" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/serialization.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1215,-598 1215,-617 1388,-617 1388,-598 1215,-598"/>
+<text text-anchor="middle" x="1301.5" y="-605" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/serialization.h</text>
</a>
</g>
</g>
-<!-- Node4->Node179 -->
+<!-- Node4->Node180 -->
<g id="edge43" class="edge">
-<title>Node4->Node179</title>
-<path fill="none" stroke="#191970" d="M1120.1987,-727.2171C1156.0397,-718.5733 1200.2455,-706.0907 1238,-690 1286.7914,-669.2055 1339.3025,-634.148 1363.4467,-617.2219"/>
-<polygon fill="#191970" stroke="#191970" points="1119.2495,-723.845 1110.3238,-729.5531 1120.861,-730.657 1119.2495,-723.845"/>
+<title>Node4->Node180</title>
+<path fill="none" stroke="#191970" d="M1123.8853,-720.8351C1169.9695,-691.429 1252.6258,-638.6864 1286.4487,-617.1042"/>
+<polygon fill="#191970" stroke="#191970" points="1121.7288,-718.0593 1115.1815,-726.389 1125.4942,-723.9603 1121.7288,-718.0593"/>
</g>
<!-- Node24 -->
<g id="node27" class="node">
<title>Node24</title>
<g id="a_node27"><a xlink:href="relay_2qnn_2transform_8h.html" target="_top" xlink:title="include/tvm/relay/qnn\l/transform.h">
-<polygon fill="#ffffff" stroke="#000000" points="392.5,-659.5 392.5,-689.5 515.5,-689.5 515.5,-659.5 392.5,-659.5"/>
-<text text-anchor="start" x="400.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/relay/qnn</text>
-<text text-anchor="middle" x="454" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/transform.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="436,-659.5 436,-689.5 559,-689.5 559,-659.5 436,-659.5"/>
+<text text-anchor="start" x="444" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/relay/qnn</text>
+<text text-anchor="middle" x="497.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/transform.h</text>
</a>
</g>
</g>
<!-- Node4->Node24 -->
<g id="edge44" class="edge">
<title>Node4->Node24</title>
-<path fill="none" stroke="#191970" d="M983.5666,-737.3906C882.7362,-730.7536 688.518,-715.7472 525,-690 522.0088,-689.529 518.9545,-689.0101 515.8751,-688.4564"/>
-<polygon fill="#191970" stroke="#191970" points="983.4135,-740.888 993.6198,-738.0449 983.8682,-733.9028 983.4135,-740.888"/>
+<path fill="none" stroke="#191970" d="M1023.1832,-737.3285C923.0049,-730.634 730.5613,-715.5793 568.5,-690 565.509,-689.5279 562.4548,-689.0081 559.3756,-688.4537"/>
+<polygon fill="#191970" stroke="#191970" points="1022.9641,-740.8215 1033.1732,-737.9887 1023.4258,-733.8368 1022.9641,-740.8215"/>
</g>
-<!-- Node180 -->
+<!-- Node181 -->
<g id="node28" class="node">
-<title>Node180</title>
+<title>Node181</title>
<g id="a_node28"><a xlink:href="builtin__fp16_8h.html" target="_top" xlink:title="Functions for conversion between fp32 and fp16. ">
-<polygon fill="#ffffff" stroke="#000000" points="534,-659.5 534,-689.5 650,-689.5 650,-659.5 534,-659.5"/>
-<text text-anchor="start" x="542" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="592" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/builtin_fp16.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="577.5,-659.5 577.5,-689.5 693.5,-689.5 693.5,-659.5 577.5,-659.5"/>
+<text text-anchor="start" x="585.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="635.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/builtin_fp16.h</text>
</a>
</g>
</g>
-<!-- Node4->Node180 -->
+<!-- Node4->Node181 -->
<g id="edge45" class="edge">
-<title>Node4->Node180</title>
-<path fill="none" stroke="#191970" d="M983.7946,-734.1216C906.1061,-725.3293 775.4861,-709.3671 664,-690 659.4648,-689.2122 654.7813,-688.3394 650.0809,-687.4211"/>
-<polygon fill="#191970" stroke="#191970" points="983.6035,-737.6221 993.9322,-735.2621 984.3862,-730.666 983.6035,-737.6221"/>
+<title>Node4->Node181</title>
+<path fill="none" stroke="#191970" d="M1023.3362,-734.0008C946.3722,-725.1526 817.5267,-709.1906 707.5,-690 702.9654,-689.2091 698.2823,-688.3341 693.5821,-687.4143"/>
+<polygon fill="#191970" stroke="#191970" points="1023.0498,-737.4908 1033.3827,-735.1492 1023.8449,-730.5361 1023.0498,-737.4908"/>
</g>
-<!-- Node181 -->
+<!-- Node182 -->
<g id="node29" class="node">
-<title>Node181</title>
+<title>Node182</title>
<g id="a_node29"><a xlink:href="c__backend__api_8h.html" target="_top" xlink:title="TVM runtime backend API. ">
-<polygon fill="#ffffff" stroke="#000000" points="258,-659.5 258,-689.5 374,-689.5 374,-659.5 258,-659.5"/>
-<text text-anchor="start" x="266" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="316" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/c_backend_api.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="263.5,-659.5 263.5,-689.5 379.5,-689.5 379.5,-659.5 263.5,-659.5"/>
+<text text-anchor="start" x="271.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="321.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/c_backend_api.h</text>
</a>
</g>
</g>
-<!-- Node4->Node181 -->
+<!-- Node4->Node182 -->
<g id="edge46" class="edge">
-<title>Node4->Node181</title>
-<path fill="none" stroke="#191970" d="M983.3306,-739.5375C861.8469,-735.3256 601.3425,-722.9655 384,-690 380.7677,-689.5097 377.4608,-688.9505 374.131,-688.3422"/>
-<polygon fill="#191970" stroke="#191970" points="983.4811,-743.0446 993.594,-739.8854 983.7183,-736.0486 983.4811,-743.0446"/>
+<title>Node4->Node182</title>
+<path fill="none" stroke="#191970" d="M1023.0757,-739.592C897.6641,-735.3989 622.788,-722.9773 393.5,-690 388.9438,-689.3447 384.2452,-688.5672 379.5343,-687.7148"/>
+<polygon fill="#191970" stroke="#191970" points="1023.1847,-743.0974 1033.2938,-739.9264 1023.4137,-736.1011 1023.1847,-743.0974"/>
</g>
-<!-- Node184 -->
+<!-- Node185 -->
<g id="node32" class="node">
-<title>Node184</title>
+<title>Node185</title>
<g id="a_node32"><a xlink:href="crt_2packed__func_8h.html" target="_top" xlink:title="Type-erased function used across TVM API. ">
-<polygon fill="#ffffff" stroke="#000000" points="172,-458.5 172,-488.5 288,-488.5 288,-458.5 172,-458.5"/>
-<text text-anchor="start" x="180" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="230" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/packed_func.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="63.5,-458.5 63.5,-488.5 179.5,-488.5 179.5,-458.5 63.5,-458.5"/>
+<text text-anchor="start" x="71.5" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="121.5" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/packed_func.h</text>
</a>
</g>
</g>
-<!-- Node4->Node184 -->
+<!-- Node4->Node185 -->
<g id="edge53" class="edge">
-<title>Node4->Node184</title>
-<path fill="none" stroke="#191970" d="M983.6033,-740.0468C798.2697,-735.7138 295.7643,-721.1999 228,-690 180.5251,-668.1417 149.7003,-642.5448 163,-592 173.6447,-551.5455 201.5986,-510.384 217.8377,-488.832"/>
-<polygon fill="#191970" stroke="#191970" points="983.6914,-743.5496 993.7696,-740.2817 983.8532,-736.5515 983.6914,-743.5496"/>
+<title>Node4->Node185</title>
+<path fill="none" stroke="#191970" d="M1023.0704,-739.3395C828.0644,-732.931 278.5149,-713.0319 200.5,-690 142.6593,-672.924 112.842,-675.12 82.5,-623 56.5892,-578.4917 91.4821,-516.6521 110.548,-488.5141"/>
+<polygon fill="#191970" stroke="#191970" points="1023.0836,-742.8417 1033.1927,-739.6705 1023.3125,-735.8454 1023.0836,-742.8417"/>
</g>
-<!-- Node185 -->
+<!-- Node186 -->
<g id="node33" class="node">
-<title>Node185</title>
-<g id="a_node33"><a xlink:href="graph__executor_8h.html" target="_top" xlink:title="Tiny graph executor that can run graph containing only tvm PackedFunc. ">
-<polygon fill="#ffffff" stroke="#000000" points="170.5,-391.5 170.5,-421.5 289.5,-421.5 289.5,-391.5 170.5,-391.5"/>
-<text text-anchor="start" x="178.5" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="230" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/graph_executor.h</text>
+<title>Node186</title>
+<g id="a_node33"><a xlink:href="graph__executor_8h.html" target="_top" xlink:title="Tiny AoT executor. ">
+<polygon fill="#ffffff" stroke="#000000" points="0,-391.5 0,-421.5 119,-421.5 119,-391.5 0,-391.5"/>
+<text text-anchor="start" x="8" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="59.5" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/graph_executor.h</text>
</a>
</g>
</g>
-<!-- Node4->Node185 -->
+<!-- Node4->Node186 -->
<g id="edge52" class="edge">
-<title>Node4->Node185</title>
-<path fill="none" stroke="#191970" d="M983.8852,-740.5633C790.5157,-737.477 247.5182,-725.697 176,-690 132.9749,-668.5248 106,-655.5869 106,-607.5 106,-607.5 106,-607.5 106,-540.5 106,-485.0315 163.9453,-442.7806 200.6137,-421.5968"/>
-<polygon fill="#191970" stroke="#191970" points="983.8707,-744.0634 993.9244,-740.7207 983.9805,-737.0642 983.8707,-744.0634"/>
+<title>Node4->Node186</title>
+<path fill="none" stroke="#191970" d="M1023.0911,-740.6797C814.1568,-737.7905 193.5075,-726.222 110.5,-690 62.2479,-668.9443 25.5,-660.146 25.5,-607.5 25.5,-607.5 25.5,-607.5 25.5,-540.5 25.5,-496.338 43.0645,-446.4343 52.9445,-421.8567"/>
+<polygon fill="#191970" stroke="#191970" points="1023.2695,-744.1824 1033.3161,-740.8187 1023.3647,-737.183 1023.2695,-744.1824"/>
</g>
-<!-- Node186 -->
+<!-- Node187 -->
<g id="node34" class="node">
-<title>Node186</title>
+<title>Node187</title>
<g id="a_node34"><a xlink:href="page__allocator_8h.html" target="_top" xlink:title="An implementation of a dynamic memory allocator for microcontrollers. ">
-<polygon fill="#ffffff" stroke="#000000" points="908,-659.5 908,-689.5 1024,-689.5 1024,-659.5 908,-659.5"/>
-<text text-anchor="start" x="916" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="966" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/page_allocator.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="928.5,-659.5 928.5,-689.5 1044.5,-689.5 1044.5,-659.5 928.5,-659.5"/>
+<text text-anchor="start" x="936.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="986.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/page_allocator.h</text>
</a>
</g>
</g>
-<!-- Node4->Node186 -->
+<!-- Node4->Node187 -->
<g id="edge54" class="edge">
-<title>Node4->Node186</title>
-<path fill="none" stroke="#191970" d="M1024.4855,-720.0643C1011.7696,-710.1577 996.9684,-698.6265 985.4375,-689.6432"/>
-<polygon fill="#191970" stroke="#191970" points="1022.4456,-722.9119 1032.4853,-726.2967 1026.7477,-717.3899 1022.4456,-722.9119"/>
+<title>Node4->Node187</title>
+<path fill="none" stroke="#191970" d="M1059.4924,-721.0761C1043.6661,-710.9774 1024.8929,-698.9983 1010.3487,-689.7177"/>
+<polygon fill="#191970" stroke="#191970" points="1057.6232,-724.0352 1067.936,-726.4639 1061.3887,-718.1342 1057.6232,-724.0352"/>
</g>
-<!-- Node187 -->
+<!-- Node188 -->
<g id="node35" class="node">
-<title>Node187</title>
+<title>Node188</title>
<g id="a_node35"><a xlink:href="platform_8h.html" target="_top" xlink:title="The virtual memory manager for micro-controllers. ">
-<polygon fill="#ffffff" stroke="#000000" points="172,-592.5 172,-622.5 288,-622.5 288,-592.5 172,-592.5"/>
-<text text-anchor="start" x="180" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="230" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/platform.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="91.5,-592.5 91.5,-622.5 207.5,-622.5 207.5,-592.5 91.5,-592.5"/>
+<text text-anchor="start" x="99.5" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="149.5" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/platform.h</text>
</a>
</g>
</g>
-<!-- Node4->Node187 -->
+<!-- Node4->Node188 -->
<g id="edge55" class="edge">
-<title>Node4->Node187</title>
-<path fill="none" stroke="#191970" d="M983.6306,-739.3218C794.5718,-733.0075 276.8172,-713.6327 249,-690 229.6218,-673.5369 227.7575,-641.6486 228.5857,-622.7016"/>
-<polygon fill="#191970" stroke="#191970" points="983.6164,-742.8231 993.7271,-739.657 983.8487,-735.827 983.6164,-742.8231"/>
+<title>Node4->Node188</title>
+<path fill="none" stroke="#191970" d="M1023.2244,-739.6316C836.3094,-734.1893 325.4425,-717.0039 254.5,-690 216.5838,-675.5674 181.4396,-642.3146 162.9958,-622.7167"/>
+<polygon fill="#191970" stroke="#191970" points="1023.3761,-743.1374 1033.4731,-739.9277 1023.5783,-736.1403 1023.3761,-743.1374"/>
</g>
-<!-- Node188 -->
+<!-- Node189 -->
<g id="node36" class="node">
-<title>Node188</title>
+<title>Node189</title>
<g id="a_node36"><a xlink:href="data__type_8h.html" target="_top" xlink:title="include/tvm/runtime\l/data_type.h">
-<polygon fill="#ffffff" stroke="#000000" points="774,-659.5 774,-689.5 890,-689.5 890,-659.5 774,-659.5"/>
-<text text-anchor="start" x="782" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="832" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/data_type.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="753.5,-659.5 753.5,-689.5 869.5,-689.5 869.5,-659.5 753.5,-659.5"/>
+<text text-anchor="start" x="761.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="811.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/data_type.h</text>
</a>
</g>
</g>
-<!-- Node4->Node188 -->
+<!-- Node4->Node189 -->
<g id="edge57" class="edge">
-<title>Node4->Node188</title>
-<path fill="none" stroke="#191970" d="M993.0545,-723.5484C958.2315,-712.9432 914.6363,-699.6665 881.5876,-689.6017"/>
-<polygon fill="#191970" stroke="#191970" points="992.0418,-726.8987 1002.6277,-726.4639 994.0812,-720.2023 992.0418,-726.8987"/>
+<title>Node4->Node189</title>
+<path fill="none" stroke="#191970" d="M1023.6924,-725.2746C976.6897,-714.0275 914.7411,-699.2041 869.9229,-688.4798"/>
+<polygon fill="#191970" stroke="#191970" points="1022.9438,-728.6942 1033.4838,-727.6176 1024.5729,-721.8864 1022.9438,-728.6942"/>
</g>
-<!-- Node191 -->
+<!-- Node192 -->
<g id="node39" class="node">
-<title>Node191</title>
+<title>Node192</title>
<g id="a_node39"><a xlink:href="ndarray_8h.html" target="_top" xlink:title="A device-independent managed NDArray abstraction. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1080,-592.5 1080,-622.5 1196,-622.5 1196,-592.5 1080,-592.5"/>
-<text text-anchor="start" x="1088" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1138" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/ndarray.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1004.5,-592.5 1004.5,-622.5 1120.5,-622.5 1120.5,-592.5 1004.5,-592.5"/>
+<text text-anchor="start" x="1012.5" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1062.5" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/ndarray.h</text>
</a>
</g>
</g>
-<!-- Node4->Node191 -->
-<g id="edge92" class="edge">
-<title>Node4->Node191</title>
-<path fill="none" stroke="#191970" d="M1067.218,-717.7882C1084.8275,-690.3502 1113.307,-645.9751 1128.2054,-622.7614"/>
-<polygon fill="#191970" stroke="#191970" points="1064.1538,-716.0827 1061.6981,-726.389 1070.0449,-719.8635 1064.1538,-716.0827"/>
+<!-- Node4->Node192 -->
+<g id="edge93" class="edge">
+<title>Node4->Node192</title>
+<path fill="none" stroke="#191970" d="M1092.3954,-715.8702C1092.3751,-699.3887 1091.2278,-677.6469 1086.5,-659 1083.2514,-646.1874 1076.828,-632.6777 1071.4435,-622.7111"/>
+<polygon fill="#191970" stroke="#191970" points="1088.8897,-716.2927 1092.2596,-726.3373 1095.8892,-716.3836 1088.8897,-716.2927"/>
</g>
-<!-- Node193 -->
+<!-- Node194 -->
<g id="node40" class="node">
-<title>Node193</title>
+<title>Node194</title>
<g id="a_node40"><a xlink:href="device__api_8h.html" target="_top" xlink:title="Abstract device memory management API. ">
-<polygon fill="#ffffff" stroke="#000000" points="1882,-458.5 1882,-488.5 1998,-488.5 1998,-458.5 1882,-458.5"/>
-<text text-anchor="start" x="1890" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1940" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/device_api.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1586.5,-458.5 1586.5,-488.5 1702.5,-488.5 1702.5,-458.5 1586.5,-458.5"/>
+<text text-anchor="start" x="1594.5" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1644.5" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/device_api.h</text>
</a>
</g>
</g>
-<!-- Node4->Node193 -->
-<g id="edge89" class="edge">
-<title>Node4->Node193</title>
-<path fill="none" stroke="#191970" d="M1120.523,-740.4253C1336.8644,-736.7011 1992.9474,-722.7415 2024,-690 2074.4641,-636.7914 2059.8254,-588.9869 2024,-525 2014.8076,-508.5816 1998.0059,-496.7643 1981.8313,-488.6253"/>
-<polygon fill="#191970" stroke="#191970" points="1120.1913,-736.9303 1110.2523,-740.6001 1120.3104,-743.9293 1120.1913,-736.9303"/>
+<!-- Node4->Node194 -->
+<g id="edge88" class="edge">
+<title>Node4->Node194</title>
+<path fill="none" stroke="#191970" d="M1159.7086,-739.7631C1366.0037,-734.2337 1971.2818,-715.8716 2004.5,-690 2039.6431,-662.6292 2051.9094,-627.8712 2025.5,-592 1986.8947,-539.5634 1799.5151,-500.2352 1702.6936,-483.0419"/>
+<polygon fill="#191970" stroke="#191970" points="1159.5097,-736.267 1149.6065,-740.0321 1159.6961,-743.2645 1159.5097,-736.267"/>
</g>
-<!-- Node194 -->
+<!-- Node195 -->
<g id="node41" class="node">
-<title>Node194</title>
+<title>Node195</title>
<g id="a_node41"><a xlink:href="profiling_8h.html" target="_top" xlink:title="Runtime profiling including timers. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1676,-391.5 1676,-421.5 1792,-421.5 1792,-391.5 1676,-391.5"/>
-<text text-anchor="start" x="1684" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1734" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/profiling.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1796.5,-391.5 1796.5,-421.5 1912.5,-421.5 1912.5,-391.5 1796.5,-391.5"/>
+<text text-anchor="start" x="1804.5" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1854.5" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/profiling.h</text>
</a>
</g>
</g>
-<!-- Node4->Node194 -->
-<g id="edge107" class="edge">
-<title>Node4->Node194</title>
-<path fill="none" stroke="#191970" d="M1120.2838,-739.4634C1266.5833,-734.5718 1605.9868,-720.2546 1717,-690 1847.7905,-654.3555 1919.225,-671.0002 1991,-556 1998.2949,-544.3119 1999.1986,-536.0729 1991,-525 1958.3724,-480.9335 1923.1571,-511.1514 1873,-489 1829.3604,-469.727 1782.3835,-439.6702 1755.7459,-421.6508"/>
-<polygon fill="#191970" stroke="#191970" points="1119.969,-735.9718 1110.0897,-739.7994 1120.1997,-742.968 1119.969,-735.9718"/>
+<!-- Node4->Node195 -->
+<g id="edge109" class="edge">
+<title>Node4->Node195</title>
+<path fill="none" stroke="#191970" d="M1159.6711,-740.348C1292.3223,-737.3087 1580.1525,-726.5833 1671.5,-690 1796.9055,-639.777 1851.2591,-615.1812 1899.5,-489 1904.4201,-476.1307 1904.3416,-470.8991 1899.5,-458 1894.1594,-443.7714 1882.9094,-430.8964 1872.9945,-421.5785"/>
+<polygon fill="#191970" stroke="#191970" points="1159.5458,-736.8498 1149.6257,-740.5702 1159.7007,-743.8481 1159.5458,-736.8498"/>
</g>
-<!-- Node196 -->
+<!-- Node197 -->
<g id="node42" class="node">
-<title>Node196</title>
+<title>Node197</title>
<g id="a_node42"><a xlink:href="packed__func_8h.html" target="_top" xlink:title="Type-erased function used across TVM API. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1672,-525.5 1672,-555.5 1788,-555.5 1788,-525.5 1672,-525.5"/>
-<text text-anchor="start" x="1680" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1730" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/packed_func.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1586.5,-525.5 1586.5,-555.5 1702.5,-555.5 1702.5,-525.5 1586.5,-525.5"/>
+<text text-anchor="start" x="1594.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1644.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/packed_func.h</text>
</a>
</g>
</g>
-<!-- Node4->Node196 -->
-<g id="edge106" class="edge">
-<title>Node4->Node196</title>
-<path fill="none" stroke="#191970" d="M1120.088,-738.5387C1264.2419,-731.8313 1590.7115,-714.1631 1635,-690 1688.4509,-660.838 1716.0495,-587.3542 1725.8114,-555.5538"/>
-<polygon fill="#191970" stroke="#191970" points="1119.8585,-735.0455 1110.0304,-739.0027 1120.1811,-742.0381 1119.8585,-735.0455"/>
+<!-- Node4->Node197 -->
+<g id="edge108" class="edge">
+<title>Node4->Node197</title>
+<path fill="none" stroke="#191970" d="M1160.0304,-739.1365C1293.5451,-733.9 1578.4022,-719.39 1611.5,-690 1650.4548,-655.4092 1649.1041,-586.1514 1646.3311,-555.5869"/>
+<polygon fill="#191970" stroke="#191970" points="1159.7682,-735.6439 1149.9106,-739.527 1160.0382,-742.6387 1159.7682,-735.6439"/>
</g>
-<!-- Node197 -->
+<!-- Node198 -->
<g id="node43" class="node">
-<title>Node197</title>
+<title>Node198</title>
<g id="a_node43"><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="#ff0000" points="2022,-458.5 2022,-488.5 2138,-488.5 2138,-458.5 2022,-458.5"/>
-<text text-anchor="start" x="2030" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="2080" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/module.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1758.5,-458.5 1758.5,-488.5 1874.5,-488.5 1874.5,-458.5 1758.5,-458.5"/>
+<text text-anchor="start" x="1766.5" y="-476.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1816.5" y="-465.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/module.h</text>
</a>
</g>
</g>
-<!-- Node4->Node197 -->
-<g id="edge91" class="edge">
-<title>Node4->Node197</title>
-<path fill="none" stroke="#191970" d="M1120.4172,-741.0114C1362.1381,-738.9544 2163.5243,-729.193 2201,-690 2261.8545,-626.3568 2148.2711,-526.2392 2100.1088,-488.5332"/>
-<polygon fill="#191970" stroke="#191970" points="1120.3221,-737.5119 1110.3517,-741.0953 1120.3805,-744.5117 1120.3221,-737.5119"/>
+<!-- Node4->Node198 -->
+<g id="edge92" class="edge">
+<title>Node4->Node198</title>
+<path fill="none" stroke="#191970" d="M1160.0683,-737.9229C1296.9097,-730.3626 1594.3344,-711.6635 1635.5,-690 1725.4616,-642.6574 1788.5015,-529.8241 1809.1588,-488.7701"/>
+<polygon fill="#191970" stroke="#191970" points="1159.501,-734.4486 1149.7076,-738.4909 1159.8842,-741.4381 1159.501,-734.4486"/>
+</g>
+<!-- Node203 -->
+<g id="node44" class="node">
+<title>Node203</title>
+<g id="a_node44"><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="1003.5,-525.5 1003.5,-555.5 1119.5,-555.5 1119.5,-525.5 1003.5,-525.5"/>
+<text text-anchor="start" x="1011.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1061.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/serializer.h</text>
+</a>
</g>
-<!-- Node202 -->
+</g>
+<!-- Node4->Node203 -->
+<g id="edge110" class="edge">
+<title>Node4->Node203</title>
+<path fill="none" stroke="#191970" d="M1084.3305,-716.7039C1078.2908,-699.0903 1068.2477,-675.6975 1053.5,-659 1033.4155,-636.2601 1010.2489,-649.5135 995.5,-623 988.8023,-610.9598 989.4564,-604.3815 995.5,-592 1003.0865,-576.4574 1017.8812,-564.2364 1031.5434,-555.6075"/>
+<polygon fill="#191970" stroke="#191970" points="1081.0288,-717.8693 1087.4293,-726.3123 1087.6909,-715.7207 1081.0288,-717.8693"/>
+</g>
+<!-- Node204 -->
<g id="node45" class="node">
-<title>Node202</title>
-<g id="a_node45"><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="1122,-525.5 1122,-555.5 1238,-555.5 1238,-525.5 1122,-525.5"/>
-<text text-anchor="start" x="1130" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1180" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/serializer.h</text>
+<title>Node204</title>
+<g id="a_node45"><a xlink:href="memory__manager_8h.html" target="_top" xlink:title="Abstract device memory management API. ">
+<polygon fill="#ffffff" stroke="#ff0000" points="1279,-525.5 1279,-555.5 1416,-555.5 1416,-525.5 1279,-525.5"/>
+<text text-anchor="start" x="1287" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1347.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/vm/memory_manager.h</text>
</a>
</g>
</g>
-<!-- Node4->Node202 -->
-<g id="edge108" class="edge">
-<title>Node4->Node202</title>
-<path fill="none" stroke="#191970" d="M1048.2915,-715.8982C1045.0785,-683.798 1044.4646,-628.5542 1071,-592 1083.3566,-574.978 1102.8102,-563.3838 1121.8841,-555.5793"/>
-<polygon fill="#191970" stroke="#191970" points="1044.8603,-716.6919 1049.4961,-726.2186 1051.8131,-715.8802 1044.8603,-716.6919"/>
+<!-- Node4->Node204 -->
+<g id="edge111" class="edge">
+<title>Node4->Node204</title>
+<path fill="none" stroke="#191970" d="M1104.376,-717.1279C1122.8564,-684.3355 1159.8711,-626.1872 1206.5,-592 1229.1986,-575.3579 1257.8402,-563.5917 1283.3706,-555.5493"/>
+<polygon fill="#191970" stroke="#191970" points="1101.2197,-715.6037 1099.4537,-726.0503 1107.3488,-718.9851 1101.2197,-715.6037"/>
</g>
-<!-- Node203 -->
-<g id="node46" class="node">
-<title>Node203</title>
-<g id="a_node46"><a xlink:href="memory__manager_8h.html" target="_top" xlink:title="Abstract device memory management API. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1844.5,-525.5 1844.5,-555.5 1981.5,-555.5 1981.5,-525.5 1844.5,-525.5"/>
-<text text-anchor="start" x="1852.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1913" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/vm/memory_manager.h</text>
+<!-- Node160 -->
+<g id="node47" class="node">
+<title>Node160</title>
+<g id="a_node47"><a xlink:href="metadata_8h.html" target="_top" xlink:title="Defines types which can be used in Metadata. ">
+<polygon fill="#ffffff" stroke="#000000" points="1900.5,-592.5 1900.5,-622.5 2016.5,-622.5 2016.5,-592.5 1900.5,-592.5"/>
+<text text-anchor="start" x="1908.5" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1958.5" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/metadata.h</text>
</a>
</g>
</g>
-<!-- Node4->Node203 -->
-<g id="edge109" class="edge">
-<title>Node4->Node203</title>
-<path fill="none" stroke="#191970" d="M1120.3469,-739.719C1258.6408,-735.4729 1566.9872,-722.5683 1667,-690 1760.8529,-659.4376 1856.31,-587.0514 1895.0359,-555.5646"/>
-<polygon fill="#191970" stroke="#191970" points="1120.18,-736.2223 1110.2898,-740.0216 1120.3906,-743.2191 1120.18,-736.2223"/>
+<!-- Node4->Node160 -->
+<g id="edge89" class="edge">
+<title>Node4->Node160</title>
+<path fill="none" stroke="#191970" d="M1159.5197,-740.1147C1363.7235,-735.6271 1959.1638,-720.0116 1987.5,-690 2005.3229,-671.1232 1986.3962,-640.6546 1971.8629,-622.5026"/>
+<polygon fill="#191970" stroke="#191970" points="1159.4383,-736.6155 1149.5168,-740.3324 1159.5907,-743.6138 1159.4383,-736.6155"/>
</g>
-<!-- Node205 -->
+<!-- Node206 -->
<g id="node48" class="node">
-<title>Node205</title>
-<g id="a_node48"><a xlink:href="metadata_8h.html" target="_top" xlink:title="Defines types which can be used in Metadata. ">
-<polygon fill="#ffffff" stroke="#000000" points="2076,-659.5 2076,-689.5 2192,-689.5 2192,-659.5 2076,-659.5"/>
-<text text-anchor="start" x="2084" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="2134" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/metadata.h</text>
+<title>Node206</title>
+<g id="a_node48"><a xlink:href="metadata__types_8h.html" target="_top" xlink:title="Defines types which can be used in metadata here which are also shared between C and C++ code bases...">
+<polygon fill="#ffffff" stroke="#ff0000" points="1862.5,-659.5 1862.5,-689.5 1978.5,-689.5 1978.5,-659.5 1862.5,-659.5"/>
+<text text-anchor="start" x="1870.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1920.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/metadata_types.h</text>
</a>
</g>
</g>
-<!-- Node4->Node205 -->
+<!-- Node4->Node206 -->
<g id="edge90" class="edge">
-<title>Node4->Node205</title>
-<path fill="none" stroke="#191970" d="M1120.4297,-739.4498C1316.7173,-733.3403 1879.2441,-714.1707 2062,-690 2066.5634,-689.3965 2071.2671,-688.6562 2075.9815,-687.8295"/>
-<polygon fill="#191970" stroke="#191970" points="1120.1362,-735.9571 1110.2494,-739.7652 1120.353,-742.9538 1120.1362,-735.9571"/>
+<title>Node4->Node206</title>
+<path fill="none" stroke="#191970" d="M1159.7807,-740.1473C1292.8919,-736.784 1596.153,-725.5572 1848.5,-690 1853.0581,-689.3577 1857.758,-688.5896 1862.4698,-687.7437"/>
+<polygon fill="#191970" stroke="#191970" points="1159.6385,-736.6496 1149.7276,-740.3944 1159.8106,-743.6475 1159.6385,-736.6496"/>
</g>
-<!-- Node206 -->
+<!-- Node208 -->
<g id="node49" class="node">
-<title>Node206</title>
+<title>Node208</title>
<g id="a_node49"><a xlink:href="object_8h.html" target="_top" xlink:title="A managed object in the TVM runtime. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1510,-659.5 1510,-689.5 1626,-689.5 1626,-659.5 1510,-659.5"/>
-<text text-anchor="start" x="1518" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="1568" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/object.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1486.5,-659.5 1486.5,-689.5 1602.5,-689.5 1602.5,-659.5 1486.5,-659.5"/>
+<text text-anchor="start" x="1494.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="1544.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/object.h</text>
</a>
</g>
</g>
-<!-- Node4->Node206 -->
-<g id="edge93" class="edge">
-<title>Node4->Node206</title>
-<path fill="none" stroke="#191970" d="M1120.402,-732.6184C1222.648,-719.3422 1413.5695,-694.552 1509.6286,-682.0792"/>
-<polygon fill="#191970" stroke="#191970" points="1119.6548,-729.1859 1110.1887,-733.9445 1120.5562,-736.1276 1119.6548,-729.1859"/>
+<!-- Node4->Node208 -->
+<g id="edge94" class="edge">
+<title>Node4->Node208</title>
+<path fill="none" stroke="#191970" d="M1159.7834,-731.4007C1248.7546,-718.2416 1402.8007,-695.4577 1486.4293,-683.0888"/>
+<polygon fill="#191970" stroke="#191970" points="1159.1747,-727.9526 1149.7944,-732.8781 1160.199,-734.8772 1159.1747,-727.9526"/>
</g>
-<!-- Node219 -->
+<!-- Node221 -->
<g id="node50" class="node">
-<title>Node219</title>
+<title>Node221</title>
<g id="a_node50"><a xlink:href="parallel__for_8h.html" target="_top" xlink:title="An implementation to run loop in parallel. ">
-<polygon fill="#ffffff" stroke="#000000" points="2248,-659.5 2248,-689.5 2362,-689.5 2362,-659.5 2248,-659.5"/>
-<text text-anchor="start" x="2256" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/support</text>
-<text text-anchor="middle" x="2305" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/parallel_for.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="2110.5,-659.5 2110.5,-689.5 2224.5,-689.5 2224.5,-659.5 2110.5,-659.5"/>
+<text text-anchor="start" x="2118.5" y="-677.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/support</text>
+<text text-anchor="middle" x="2167.5" y="-666.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/parallel_for.h</text>
</a>
</g>
</g>
-<!-- Node4->Node219 -->
-<g id="edge110" class="edge">
-<title>Node4->Node219</title>
-<path fill="none" stroke="#191970" d="M1120.2239,-740.0655C1338.2128,-735.2397 2016.3242,-718.2901 2234,-690 2238.5041,-689.4146 2243.1455,-688.6874 2247.7965,-687.8697"/>
-<polygon fill="#191970" stroke="#191970" points="1120.1155,-736.5669 1110.1949,-740.2861 1120.2695,-743.5652 1120.1155,-736.5669"/>
+<!-- Node4->Node221 -->
+<g id="edge112" class="edge">
+<title>Node4->Node221</title>
+<path fill="none" stroke="#191970" d="M1159.9375,-739.4553C1355.5743,-733.378 1914.7909,-714.3235 2096.5,-690 2101.0019,-689.3974 2105.6416,-688.6577 2110.2915,-687.8315"/>
+<polygon fill="#191970" stroke="#191970" points="1159.6766,-735.9616 1149.7895,-739.7691 1159.8929,-742.9583 1159.6766,-735.9616"/>
</g>
<!-- Node6 -->
<g id="node3" class="node">
<title>Node6</title>
<g id="a_node3"><a xlink:href="auto__scheduler_2cost__model_8h.html" target="_top" xlink:title="Cost models that estimate the performance of programs. ">
-<polygon fill="#ffffff" stroke="#000000" points="630,-324.5 630,-354.5 782,-354.5 782,-324.5 630,-324.5"/>
-<text text-anchor="start" x="638" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="706" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/cost_model.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="942.5,-324.5 942.5,-354.5 1094.5,-354.5 1094.5,-324.5 942.5,-324.5"/>
+<text text-anchor="start" x="950.5" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="1018.5" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/cost_model.h</text>
</a>
</g>
</g>
<!-- Node5->Node6 -->
<g id="edge2" class="edge">
<title>Node5->Node6</title>
-<path fill="none" stroke="#191970" d="M162.4391,-390.8691C335.0971,-367.378 526.3897,-352.0273 629.9377,-344.6087"/>
-<polygon fill="#191970" stroke="#191970" points="161.8368,-387.4189 152.4057,-392.2464 162.7888,-394.3539 161.8368,-387.4189"/>
+<path fill="none" stroke="#191970" d="M428.0413,-399.336C537.1364,-390.1023 730.3876,-373.073 895.5,-355 910.6906,-353.3373 926.9163,-351.4046 942.4345,-349.4795"/>
+<polygon fill="#191970" stroke="#191970" points="427.4202,-395.8759 417.7502,-400.205 428.0093,-402.851 427.4202,-395.8759"/>
</g>
<!-- Node7 -->
<g id="node4" class="node">
<title>Node7</title>
<g id="a_node4"><a xlink:href="auto__scheduler_2feature_8h.html" target="_top" xlink:title="Feature extraction for the cost model. We extract one feature vector per BufferStoreNode statement in...">
-<polygon fill="#ffffff" stroke="#000000" points="0,-324.5 0,-354.5 152,-354.5 152,-324.5 0,-324.5"/>
-<text text-anchor="start" x="8" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="76" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/feature.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="216.5,-324.5 216.5,-354.5 368.5,-354.5 368.5,-324.5 216.5,-324.5"/>
+<text text-anchor="start" x="224.5" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="292.5" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/feature.h</text>
</a>
</g>
</g>
<!-- Node5->Node7 -->
<g id="edge3" class="edge">
<title>Node5->Node7</title>
-<path fill="none" stroke="#191970" d="M76,-381.0249C76,-372.128 76,-362.4287 76,-354.6432"/>
-<polygon fill="#191970" stroke="#191970" points="72.5001,-381.2966 76,-391.2967 79.5001,-381.2967 72.5001,-381.2966"/>
+<path fill="none" stroke="#191970" d="M324.3592,-383.0626C317.4619,-373.6315 309.7068,-363.0276 303.5749,-354.6432"/>
+<polygon fill="#191970" stroke="#191970" points="321.6528,-385.2911 330.3811,-391.2967 327.303,-381.1588 321.6528,-385.2911"/>
</g>
<!-- Node8 -->
<g id="node5" class="node">
<title>Node8</title>
<g id="a_node5"><a xlink:href="search__task_8h.html" target="_top" xlink:title="Meta information and hardware parameters for a search task. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="533,-67.5 533,-97.5 685,-97.5 685,-67.5 533,-67.5"/>
-<text text-anchor="start" x="541" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="609" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/search_task.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="386.5,-67.5 386.5,-97.5 538.5,-97.5 538.5,-67.5 386.5,-67.5"/>
+<text text-anchor="start" x="394.5" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="462.5" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/search_task.h</text>
</a>
</g>
</g>
<!-- Node5->Node8 -->
<g id="edge4" class="edge">
<title>Node5->Node8</title>
-<path fill="none" stroke="#191970" d="M130.6044,-387.9858C201.1484,-362.3726 315,-314.5254 315,-272.5 315,-272.5 315,-272.5 315,-211 315,-161.7019 455.5445,-119.1453 542.2868,-97.5746"/>
-<polygon fill="#191970" stroke="#191970" points="129.1838,-384.777 120.9541,-391.4495 131.5486,-391.3655 129.1838,-384.777"/>
+<path fill="none" stroke="#191970" d="M360.7564,-383.019C366.869,-374.5998 373.1596,-364.7662 377.5,-355 392.7813,-320.6163 396.5,-310.1265 396.5,-272.5 396.5,-272.5 396.5,-272.5 396.5,-211 396.5,-165.8567 429.1294,-120.6677 448.5031,-97.8155"/>
+<polygon fill="#191970" stroke="#191970" points="357.7925,-381.1351 354.5403,-391.2185 363.3707,-385.3641 357.7925,-381.1351"/>
</g>
<!-- Node12 -->
<g id="node6" class="node">
<title>Node12</title>
<g id="a_node6"><a xlink:href="search__policy_8h.html" target="_top" xlink:title="The base class of search policies, including the abstract definition of search policy and other suppo...">
-<polygon fill="#ffffff" stroke="#ff0000" points="593,-.5 593,-30.5 745,-30.5 745,-.5 593,-.5"/>
-<text text-anchor="start" x="601" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="669" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/search_policy.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="334.5,-.5 334.5,-30.5 486.5,-30.5 486.5,-.5 334.5,-.5"/>
+<text text-anchor="start" x="342.5" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="410.5" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/search_policy.h</text>
</a>
</g>
</g>
<!-- Node8->Node12 -->
<g id="edge5" class="edge">
<title>Node8->Node12</title>
-<path fill="none" stroke="#191970" d="M629.473,-59.6385C638.046,-50.0653 647.7773,-39.1987 655.439,-30.6432"/>
-<polygon fill="#191970" stroke="#191970" points="626.6788,-57.5122 622.6149,-67.2967 631.8935,-62.1821 626.6788,-57.5122"/>
+<path fill="none" stroke="#191970" d="M444.5335,-59.3509C437.1583,-49.8482 428.8266,-39.1132 422.2529,-30.6432"/>
+<polygon fill="#191970" stroke="#191970" points="441.8042,-61.5427 450.7004,-67.2967 447.3341,-57.2508 441.8042,-61.5427"/>
</g>
<!-- Node13->Node6 -->
<g id="edge7" class="edge">
<title>Node13->Node6</title>
-<path fill="none" stroke="#191970" d="M1121.7899,-395.1303C1113.4839,-393.6535 1105.0353,-392.2294 1097,-391 963.8187,-370.6225 929.3213,-374.44 796,-355 791.4502,-354.3366 786.7762,-353.6245 782.0646,-352.8835"/>
-<polygon fill="#191970" stroke="#191970" points="1121.349,-398.6073 1131.8131,-396.9481 1122.5982,-391.7197 1121.349,-398.6073"/>
+<path fill="none" stroke="#191970" d="M865.4776,-393.4605C897.1342,-382.2974 942.7224,-366.2216 975.901,-354.5218"/>
+<polygon fill="#191970" stroke="#191970" points="863.9892,-390.2741 855.7224,-396.9005 866.3172,-396.8757 863.9892,-390.2741"/>
</g>
<!-- Node13->Node12 -->
<g id="edge8" class="edge">
<title>Node13->Node12</title>
-<path fill="none" stroke="#191970" d="M1120.8648,-395.0893C1112.8544,-393.6512 1104.7332,-392.2486 1097,-391 980.0947,-372.125 942.7203,-399.5484 833,-355 811.5036,-346.2721 809.6818,-337.7572 791,-324 769.2399,-307.976 756.7081,-310.6703 742,-288 687.2942,-203.6792 726.846,-161.994 694,-67 689.6479,-54.4132 683.0022,-40.8929 677.6565,-30.865"/>
-<polygon fill="#191970" stroke="#191970" points="1120.4488,-398.5711 1130.9143,-396.9208 1121.7039,-391.6845 1120.4488,-398.5711"/>
+<path fill="none" stroke="#191970" d="M749.4309,-398.6945C670.6079,-389.9917 555.6687,-374.5354 515.5,-355 430.3523,-313.5897 358.5,-305.6834 358.5,-211 358.5,-211 358.5,-211 358.5,-149.5 358.5,-111.8735 362.9131,-101.6839 377.5,-67 383.0165,-53.8833 391.7994,-40.3878 398.9179,-30.5039"/>
+<polygon fill="#191970" stroke="#191970" points="749.4993,-402.2226 759.8194,-399.8256 750.257,-395.2637 749.4993,-402.2226"/>
</g>
<!-- Node14 -->
<g id="node8" class="node">
<title>Node14</title>
<g id="a_node8"><a xlink:href="transform__step_8h.html" target="_top" xlink:title="Transformation steps. These steps are used to manipulate LoopState. They are similar to the schedule ...">
-<polygon fill="#ffffff" stroke="#ff0000" points="990,-324.5 990,-354.5 1142,-354.5 1142,-324.5 990,-324.5"/>
-<text text-anchor="start" x="998" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
-<text text-anchor="middle" x="1066" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/transform_step.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="734.5,-324.5 734.5,-354.5 886.5,-354.5 886.5,-324.5 734.5,-324.5"/>
+<text text-anchor="start" x="742.5" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/auto_scheduler</text>
+<text text-anchor="middle" x="810.5" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/transform_step.h</text>
</a>
</g>
</g>
<!-- Node13->Node14 -->
<g id="edge9" class="edge">
<title>Node13->Node14</title>
-<path fill="none" stroke="#191970" d="M1155.0449,-391.8334C1136.2313,-380.7763 1110.5186,-365.6644 1091.5594,-354.5218"/>
-<polygon fill="#191970" stroke="#191970" points="1153.2718,-394.851 1163.6666,-396.9005 1156.8187,-388.8161 1153.2718,-394.851"/>
+<path fill="none" stroke="#191970" d="M823.2892,-387.1042C820.5017,-376.7287 817.121,-364.145 814.5357,-354.5218"/>
+<polygon fill="#191970" stroke="#191970" points="819.9463,-388.1511 825.921,-396.9005 826.7065,-386.3349 819.9463,-388.1511"/>
</g>
<!-- Node16 -->
<g id="node9" class="node">
<title>Node16</title>
<g id="a_node9"><a xlink:href="ir_2adt_8h.html" target="_top" xlink:title="Algebraic data type definitions. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1407,-140 1407,-159 1519,-159 1519,-140 1407,-140"/>
-<text text-anchor="middle" x="1463" y="-147" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/adt.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1302.5,-140 1302.5,-159 1414.5,-159 1414.5,-140 1302.5,-140"/>
+<text text-anchor="middle" x="1358.5" y="-147" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/adt.h</text>
</a>
</g>
</g>
<!-- Node13->Node16 -->
<g id="edge10" class="edge">
<title>Node13->Node16</title>
-<path fill="none" stroke="#191970" d="M1258.9261,-396.0383C1312.3906,-387.4449 1376.8402,-373.6503 1397,-355 1412.1973,-340.9406 1450.3423,-198.0058 1460.4885,-159.1849"/>
-<polygon fill="#191970" stroke="#191970" points="1258.0779,-392.6284 1248.7406,-397.6349 1259.162,-399.544 1258.0779,-392.6284"/>
+<path fill="none" stroke="#191970" d="M885.7688,-395.1443C894.3728,-393.6302 903.1559,-392.1918 911.5,-391 991.5291,-379.5694 1208.6026,-403.2041 1273.5,-355 1340.4471,-305.2734 1354.9337,-193.1621 1357.8263,-159.3413"/>
+<polygon fill="#191970" stroke="#191970" points="885.007,-391.7251 875.788,-396.9462 886.2507,-398.6138 885.007,-391.7251"/>
</g>
<!-- Node135 -->
<g id="node10" class="node">
<title>Node135</title>
<g id="a_node10"><a xlink:href="ir_2expr_8h.html" target="_top" xlink:title="Base expr nodes in TVM. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1270.5,-201.5 1270.5,-220.5 1387.5,-220.5 1387.5,-201.5 1270.5,-201.5"/>
-<text text-anchor="middle" x="1329" y="-208.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/expr.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1386,-201.5 1386,-220.5 1503,-220.5 1503,-201.5 1386,-201.5"/>
+<text text-anchor="middle" x="1444.5" y="-208.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/expr.h</text>
</a>
</g>
</g>
<!-- Node13->Node135 -->
<g id="edge11" class="edge">
<title>Node13->Node135</title>
-<path fill="none" stroke="#191970" d="M1189.1015,-387.9374C1197.8688,-370.7429 1211.9818,-344.7693 1227,-324 1256.8876,-282.6672 1299.6845,-239.4302 1318.9386,-220.6527"/>
-<polygon fill="#191970" stroke="#191970" points="1185.9407,-386.4327 1184.5907,-396.9412 1192.1993,-389.5682 1185.9407,-386.4327"/>
+<path fill="none" stroke="#191970" d="M885.4375,-395.1388C894.1455,-393.6124 903.0467,-392.1719 911.5,-391 998.8066,-378.8965 1230.0405,-398.8507 1306.5,-355 1352.9314,-328.3709 1338.1724,-294.3629 1376.5,-257 1391.707,-242.1758 1412.2478,-229.0249 1426.762,-220.6213"/>
+<polygon fill="#191970" stroke="#191970" points="884.5624,-391.74 875.3422,-396.9591 885.8045,-398.629 884.5624,-391.74"/>
</g>
<!-- Node26 -->
<g id="node11" class="node">
<title>Node26</title>
<g id="a_node11"><a xlink:href="target_8h.html" target="_top" xlink:title="Compilation target object. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="788.5,-134.5 788.5,-164.5 895.5,-164.5 895.5,-134.5 788.5,-134.5"/>
-<text text-anchor="start" x="796.5" y="-152.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
-<text text-anchor="middle" x="842" y="-141.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/target.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="577,-134.5 577,-164.5 684,-164.5 684,-134.5 577,-134.5"/>
+<text text-anchor="start" x="585" y="-152.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
+<text text-anchor="middle" x="630.5" y="-141.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/target.h</text>
</a>
</g>
</g>
<!-- Node13->Node26 -->
<g id="edge33" class="edge">
<title>Node13->Node26</title>
-<path fill="none" stroke="#191970" d="M1119.2479,-395.1059C1039.9347,-380.2035 910.9933,-355.8709 910,-355 895.0596,-341.9 858.3918,-210.1796 846.0955,-164.7821"/>
-<polygon fill="#191970" stroke="#191970" points="1118.6641,-398.5574 1129.1383,-396.9637 1119.9564,-391.6777 1118.6641,-398.5574"/>
+<path fill="none" stroke="#191970" d="M763.3811,-394.9494C720.1456,-385.8926 668.8038,-371.9942 654.5,-355 607.3141,-298.9389 620.0905,-201.7307 627.2278,-164.5529"/>
+<polygon fill="#191970" stroke="#191970" points="762.8178,-398.4065 773.3157,-396.9768 764.2175,-391.5478 762.8178,-398.4065"/>
</g>
<!-- Node61 -->
<g id="node12" class="node">
<title>Node61</title>
<g id="a_node12"><a xlink:href="tag_8h.html" target="_top" xlink:title="Target tag registry. ">
-<polygon fill="#ffffff" stroke="#000000" points="835.5,-67.5 835.5,-97.5 942.5,-97.5 942.5,-67.5 835.5,-67.5"/>
-<text text-anchor="start" x="843.5" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
-<text text-anchor="middle" x="889" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/tag.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="627,-67.5 627,-97.5 734,-97.5 734,-67.5 627,-67.5"/>
+<text text-anchor="start" x="635" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
+<text text-anchor="middle" x="680.5" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/tag.h</text>
</a>
</g>
</g>
<!-- Node13->Node61 -->
<g id="edge32" class="edge">
<title>Node13->Node61</title>
-<path fill="none" stroke="#191970" d="M1119.0742,-395.0446C1058.6294,-383.2247 973.2691,-365.1279 961,-355 930.0093,-329.4179 924,-312.6854 924,-272.5 924,-272.5 924,-272.5 924,-211 924,-168.6872 906.1149,-121.2048 895.9149,-97.5656"/>
-<polygon fill="#191970" stroke="#191970" points="1118.4365,-398.486 1128.9206,-396.9578 1119.7718,-391.6145 1118.4365,-398.486"/>
+<path fill="none" stroke="#191970" d="M773.2129,-393.6878C752.1627,-385.9789 730.0125,-373.8575 716.5,-355 658.9805,-274.7279 710.7404,-231.2361 693.5,-134 691.3106,-121.6515 687.7784,-107.8886 684.9482,-97.6992"/>
+<polygon fill="#191970" stroke="#191970" points="772.178,-397.0325 782.7723,-396.9275 774.4249,-390.4029 772.178,-397.0325"/>
</g>
-<!-- Node13->Node162 -->
+<!-- Node13->Node163 -->
<g id="edge36" class="edge">
-<title>Node13->Node162</title>
-<path fill="none" stroke="#191970" d="M1120.2725,-394.8748C1064.881,-383.4896 989.7147,-366.2571 981,-355 916.0934,-271.1577 1001.8869,-130.1712 1027.2998,-92.1767"/>
-<polygon fill="#191970" stroke="#191970" points="1119.8675,-398.364 1130.3648,-396.9298 1121.2643,-391.5048 1119.8675,-398.364"/>
+<title>Node13->Node163</title>
+<path fill="none" stroke="#191970" d="M777.2581,-393.5171C757.5436,-385.7517 737.0167,-373.6343 725.5,-355 718.2566,-343.2799 721.5964,-337.2132 725.5,-324 754.0012,-227.5273 780.5698,-208.1943 848.5,-134 863.9088,-117.1703 885.2413,-101.6134 899.4362,-92.098"/>
+<polygon fill="#191970" stroke="#191970" points="776.2337,-396.8686 786.8278,-396.9891 778.6211,-390.2883 776.2337,-396.8686"/>
</g>
-<!-- Node163 -->
+<!-- Node164 -->
<g id="node16" class="node">
-<title>Node163</title>
+<title>Node164</title>
<g id="a_node16"><a xlink:href="relay_2base_8h.html" target="_top" xlink:title="Base classes for the Relay IR. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1608.5,-6 1608.5,-25 1745.5,-25 1745.5,-6 1608.5,-6"/>
-<text text-anchor="middle" x="1677" y="-13" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/relay/base.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="846,-6 846,-25 983,-25 983,-6 846,-6"/>
+<text text-anchor="middle" x="914.5" y="-13" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/relay/base.h</text>
</a>
</g>
</g>
-<!-- Node13->Node163 -->
+<!-- Node13->Node164 -->
<g id="edge31" class="edge">
-<title>Node13->Node163</title>
-<path fill="none" stroke="#191970" d="M1258.8622,-402.7711C1392.656,-395.8177 1652.9909,-379.3376 1687,-355 1744.2523,-314.0291 1753,-281.402 1753,-211 1753,-211 1753,-211 1753,-149.5 1753,-96.8121 1707.9944,-45.8623 1687.2497,-25.1761"/>
-<polygon fill="#191970" stroke="#191970" points="1258.4764,-399.2862 1248.6693,-403.2951 1258.8358,-406.277 1258.4764,-399.2862"/>
+<title>Node13->Node164</title>
+<path fill="none" stroke="#191970" d="M749.8624,-399.0312C699.9231,-392.1578 634.8789,-379.1858 581.5,-355 527.9081,-330.7176 476.425,-313.5979 492.5,-257 520.1837,-159.5296 532.1692,-120.0457 618.5,-67 655.6645,-44.1644 770.6093,-29.1851 845.8406,-21.5436"/>
+<polygon fill="#191970" stroke="#191970" points="749.5875,-402.5253 759.9601,-400.3667 750.5053,-395.5857 749.5875,-402.5253"/>
</g>
-<!-- Node169 -->
+<!-- Node170 -->
<g id="node17" class="node">
-<title>Node169</title>
+<title>Node170</title>
<g id="a_node17"><a xlink:href="var_8h.html" target="_top" xlink:title="Variables in the TIR. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1070.5,-140 1070.5,-159 1185.5,-159 1185.5,-140 1070.5,-140"/>
-<text text-anchor="middle" x="1128" y="-147" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/var.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="857,-140 857,-159 972,-159 972,-140 857,-140"/>
+<text text-anchor="middle" x="914.5" y="-147" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/var.h</text>
</a>
</g>
</g>
-<!-- Node13->Node169 -->
+<!-- Node13->Node170 -->
<g id="edge37" class="edge">
-<title>Node13->Node169</title>
-<path fill="none" stroke="#191970" d="M1176.0236,-386.8475C1165.5809,-335.2362 1137.6542,-197.2143 1129.9769,-159.2706"/>
-<polygon fill="#191970" stroke="#191970" points="1172.6113,-387.6319 1178.025,-396.7391 1179.4723,-386.2436 1172.6113,-387.6319"/>
+<title>Node13->Node170</title>
+<path fill="none" stroke="#191970" d="M857.4553,-391.6972C871.3891,-382.9847 886.9008,-370.5984 895.5,-355 932.3441,-288.1667 921.0306,-190.2855 916.1904,-159.2891"/>
+<polygon fill="#191970" stroke="#191970" points="855.4656,-388.8067 848.6152,-396.889 859.0106,-394.8428 855.4656,-388.8067"/>
</g>
-<!-- Node171 -->
+<!-- Node172 -->
<g id="node18" class="node">
-<title>Node171</title>
+<title>Node172</title>
<g id="a_node18"><a xlink:href="ir_2span_8h.html" target="_top" xlink:title="Span information for debugging purposes. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1558.5,-330 1558.5,-349 1677.5,-349 1677.5,-330 1558.5,-330"/>
-<text text-anchor="middle" x="1618" y="-337" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/span.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1511,-330 1511,-349 1630,-349 1630,-330 1511,-330"/>
+<text text-anchor="middle" x="1570.5" y="-337" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/span.h</text>
</a>
</g>
</g>
-<!-- Node13->Node171 -->
+<!-- Node13->Node172 -->
<g id="edge23" class="edge">
-<title>Node13->Node171</title>
-<path fill="none" stroke="#191970" d="M1252.185,-395.458C1337.7485,-382.3695 1478.7381,-360.8026 1558.383,-348.6195"/>
-<polygon fill="#191970" stroke="#191970" points="1251.6453,-391.9998 1242.2896,-396.9717 1252.7038,-398.9193 1251.6453,-391.9998"/>
+<title>Node13->Node172</title>
+<path fill="none" stroke="#191970" d="M905.8746,-395.7C919.0818,-394.014 932.6746,-392.377 945.5,-391 1139.2666,-370.1967 1188.2783,-371.0065 1382.5,-355 1425.4203,-351.4628 1473.8521,-347.4697 1510.8084,-344.4224"/>
+<polygon fill="#191970" stroke="#191970" points="905.337,-392.2403 895.8692,-396.9953 906.2358,-399.1824 905.337,-392.2403"/>
</g>
-<!-- Node172 -->
+<!-- Node173 -->
<g id="node19" class="node">
-<title>Node172</title>
+<title>Node173</title>
<g id="a_node19"><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="#ff0000" points="1444,-263 1444,-282 1562,-282 1562,-263 1444,-263"/>
-<text text-anchor="middle" x="1503" y="-270" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/type.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1385.5,-263 1385.5,-282 1503.5,-282 1503.5,-263 1385.5,-263"/>
+<text text-anchor="middle" x="1444.5" y="-270" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/type.h</text>
</a>
</g>
</g>
-<!-- Node13->Node172 -->
+<!-- Node13->Node173 -->
<g id="edge29" class="edge">
-<title>Node13->Node172</title>
-<path fill="none" stroke="#191970" d="M1258.709,-395.8255C1320.7077,-386.4054 1401.3006,-371.6151 1430,-355 1461.6131,-336.698 1486.6894,-299.7819 1497.3854,-282.2154"/>
-<polygon fill="#191970" stroke="#191970" points="1257.9111,-392.4059 1248.5396,-397.3482 1258.9477,-399.3287 1257.9111,-392.4059"/>
+<title>Node13->Node173</title>
+<path fill="none" stroke="#191970" d="M884.8196,-395.1852C893.7224,-393.6271 902.8446,-392.1649 911.5,-391 1007.1914,-378.121 1255.0637,-391.3848 1344.5,-355 1384.7358,-338.6311 1421.141,-300.0525 1436.6225,-282.056"/>
+<polygon fill="#191970" stroke="#191970" points="884.1179,-391.7552 874.8977,-396.9742 885.3601,-398.6441 884.1179,-391.7552"/>
</g>
<!-- Node144 -->
<g id="node20" class="node">
<title>Node144</title>
<g id="a_node20"><a xlink:href="arg__info_8h.html" target="_top" xlink:title="include/tvm/meta_schedule\l/arg_info.h">
-<polygon fill="#ffffff" stroke="#ff0000" points="1236,-324.5 1236,-354.5 1388,-354.5 1388,-324.5 1236,-324.5"/>
-<text text-anchor="start" x="1244" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/meta_schedule</text>
-<text text-anchor="middle" x="1312" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/arg_info.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1112.5,-324.5 1112.5,-354.5 1264.5,-354.5 1264.5,-324.5 1112.5,-324.5"/>
+<text text-anchor="start" x="1120.5" y="-342.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/meta_schedule</text>
+<text text-anchor="middle" x="1188.5" y="-331.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/arg_info.h</text>
</a>
</g>
</g>
<!-- Node13->Node144 -->
<g id="edge30" class="edge">
<title>Node13->Node144</title>
-<path fill="none" stroke="#191970" d="M1208.0087,-392.2834C1229.8661,-381.1892 1260.1528,-365.8164 1282.4049,-354.5218"/>
-<polygon fill="#191970" stroke="#191970" points="1206.2453,-389.2534 1198.9124,-396.9005 1209.4136,-395.4954 1206.2453,-389.2534"/>
+<path fill="none" stroke="#191970" d="M889.4661,-395.109C896.8887,-393.7239 904.3667,-392.3292 911.5,-391 979.4225,-378.3438 1056.7204,-363.9754 1112.3697,-353.6373"/>
+<polygon fill="#191970" stroke="#191970" points="888.6043,-391.7093 879.4163,-396.9849 889.8888,-398.5905 888.6043,-391.7093"/>
</g>
<!-- Node25 -->
<g id="node21" class="node">
<title>Node25</title>
<g id="a_node21"><a xlink:href="target__kind_8h.html" target="_top" xlink:title="Target kind registry. ">
-<polygon fill="#ffffff" stroke="#000000" points="750.5,-257.5 750.5,-287.5 857.5,-287.5 857.5,-257.5 750.5,-257.5"/>
-<text text-anchor="start" x="758.5" y="-275.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
-<text text-anchor="middle" x="804" y="-264.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/target_kind.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="501,-257.5 501,-287.5 608,-287.5 608,-257.5 501,-257.5"/>
+<text text-anchor="start" x="509" y="-275.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/target</text>
+<text text-anchor="middle" x="554.5" y="-264.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/target_kind.h</text>
</a>
</g>
</g>
<!-- Node13->Node25 -->
<g id="edge34" class="edge">
<title>Node13->Node25</title>
-<path fill="none" stroke="#191970" d="M1119.9444,-395.094C1112.2288,-393.6847 1104.4316,-392.2882 1097,-391 997.2161,-373.7031 959.8831,-405.3245 872,-355 843.9734,-338.9511 822.6311,-306.6057 811.8343,-287.5311"/>
-<polygon fill="#191970" stroke="#191970" points="1119.5342,-398.5772 1130.0027,-396.9459 1120.8018,-391.693 1119.5342,-398.5772"/>
+<path fill="none" stroke="#191970" d="M760.6007,-395.1443C707.9296,-385.3994 640.0518,-370.5085 616.5,-355 590.6126,-337.9535 571.3994,-306.2703 561.6622,-287.5176"/>
+<polygon fill="#191970" stroke="#191970" points="760.1135,-398.6131 770.579,-396.9631 761.3688,-391.7265 760.1135,-398.6131"/>
</g>
<!-- Node135->Node16 -->
<g id="edge12" class="edge">
<title>Node135->Node16</title>
-<path fill="none" stroke="#191970" d="M1359.0911,-197.1895C1384.4082,-185.5701 1419.9454,-169.2601 1442.2095,-159.0419"/>
-<polygon fill="#191970" stroke="#191970" points="1357.5661,-194.0384 1349.9375,-201.3906 1360.486,-200.4003 1357.5661,-194.0384"/>
+<path fill="none" stroke="#191970" d="M1422.7812,-195.4685C1406.7491,-184.0037 1385.4396,-168.765 1371.8431,-159.0419"/>
+<polygon fill="#191970" stroke="#191970" points="1420.8924,-198.4207 1431.0625,-201.3906 1424.9643,-192.7268 1420.8924,-198.4207"/>
</g>
<!-- Node135->Node26 -->
<g id="edge13" class="edge">
<title>Node135->Node26</title>
-<path fill="none" stroke="#191970" d="M1260.1628,-202.307C1162.2021,-189.9362 984.6236,-167.511 895.7278,-156.2849"/>
-<polygon fill="#191970" stroke="#191970" points="1259.8967,-205.8011 1270.2564,-203.5817 1260.7738,-198.8563 1259.8967,-205.8011"/>
+<path fill="none" stroke="#191970" d="M1375.8407,-205.8126C1217.0739,-193.8173 827.2305,-164.3635 684.2236,-153.559"/>
+<polygon fill="#191970" stroke="#191970" points="1375.7238,-209.3137 1385.9591,-206.5771 1376.2513,-202.3335 1375.7238,-209.3137"/>
</g>
<!-- Node28 -->
<g id="node13" class="node">
<title>Node28</title>
<g id="a_node13"><a xlink:href="tir_2usmp_2utils_8h.html" target="_top" xlink:title="Utilities for Unified Static Memory Planner. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="1409,-67.5 1409,-97.5 1527,-97.5 1527,-67.5 1409,-67.5"/>
-<text text-anchor="start" x="1417" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/usmp</text>
-<text text-anchor="middle" x="1468" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/utils.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1545.5,-67.5 1545.5,-97.5 1663.5,-97.5 1663.5,-67.5 1545.5,-67.5"/>
+<text text-anchor="start" x="1553.5" y="-85.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/usmp</text>
+<text text-anchor="middle" x="1604.5" y="-74.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/utils.h</text>
</a>
</g>
</g>
<!-- Node135->Node28 -->
<g id="edge20" class="edge">
<title>Node135->Node28</title>
-<path fill="none" stroke="#191970" d="M1343.0628,-193.3133C1356.3842,-177.064 1377.3685,-152.7334 1398,-134 1412.7442,-120.6123 1430.8943,-107.3384 1445.0179,-97.6203"/>
-<polygon fill="#191970" stroke="#191970" points="1340.0707,-191.4469 1336.5013,-201.4224 1345.5124,-195.8501 1340.0707,-191.4469"/>
+<path fill="none" stroke="#191970" d="M1464.4958,-194.9409C1495.5958,-169.9637 1555.1375,-122.1442 1585.5719,-97.7016"/>
+<polygon fill="#191970" stroke="#191970" points="1462.1312,-192.3509 1456.526,-201.3416 1466.5145,-197.8087 1462.1312,-192.3509"/>
</g>
-<!-- Node135->Node162 -->
+<!-- Node135->Node163 -->
<g id="edge18" class="edge">
-<title>Node135->Node162</title>
-<path fill="none" stroke="#191970" d="M1306.2487,-195.9241C1280.1987,-179.1441 1235.8428,-152.0228 1195,-134 1154.0557,-115.9324 1104.7598,-101.1153 1071.5922,-92.0875"/>
-<polygon fill="#191970" stroke="#191970" points="1304.3723,-198.8788 1314.6637,-201.3962 1308.1884,-193.0105 1304.3723,-198.8788"/>
+<title>Node135->Node163</title>
+<path fill="none" stroke="#191970" d="M1444.938,-191.2498C1444.173,-173.434 1439.9716,-147.9498 1423.5,-134 1389.9073,-105.5504 1103.0879,-90.2789 975.3925,-84.8366"/>
+<polygon fill="#191970" stroke="#191970" points="1441.4398,-191.4531 1445.0443,-201.4159 1448.4394,-191.3798 1441.4398,-191.4531"/>
</g>
-<!-- Node135->Node169 -->
+<!-- Node135->Node170 -->
<g id="edge21" class="edge">
-<title>Node135->Node169</title>
-<path fill="none" stroke="#191970" d="M1287.8197,-198.4001C1249.6697,-186.7273 1193.7588,-169.6202 1159.1857,-159.0419"/>
-<polygon fill="#191970" stroke="#191970" points="1287.0073,-201.8116 1297.5938,-201.3906 1289.0554,-195.1179 1287.0073,-201.8116"/>
+<title>Node135->Node170</title>
+<path fill="none" stroke="#191970" d="M1375.5401,-202.9981C1270.0797,-190.7607 1070.6376,-167.6179 972.2974,-156.2067"/>
+<polygon fill="#191970" stroke="#191970" points="1375.4189,-206.5074 1385.7557,-204.1834 1376.2258,-199.554 1375.4189,-206.5074"/>
</g>
<!-- Node26->Node8 -->
<g id="edge14" class="edge">
<title>Node26->Node8</title>
-<path fill="none" stroke="#191970" d="M780.0437,-131.6842C743.0855,-121.0568 696.6689,-107.7095 661.5178,-97.6017"/>
-<polygon fill="#191970" stroke="#191970" points="779.1324,-135.064 789.7103,-134.4639 781.067,-128.3366 779.1324,-135.064"/>
+<path fill="none" stroke="#191970" d="M583.4286,-130.7275C557.1931,-120.2645 524.957,-107.4084 500.3669,-97.6017"/>
+<polygon fill="#191970" stroke="#191970" points="582.2124,-134.0105 592.7975,-134.4639 584.8055,-127.5085 582.2124,-134.0105"/>
</g>
<!-- Node26->Node61 -->
<g id="edge15" class="edge">
<title>Node26->Node61</title>
-<path fill="none" stroke="#191970" d="M858.4412,-126.0626C865.057,-116.6315 872.4955,-106.0276 878.3772,-97.6432"/>
-<polygon fill="#191970" stroke="#191970" points="855.5426,-124.1001 852.665,-134.2967 861.2732,-128.1201 855.5426,-124.1001"/>
+<path fill="none" stroke="#191970" d="M647.9906,-126.0626C655.0287,-116.6315 662.9421,-106.0276 669.1991,-97.6432"/>
+<polygon fill="#191970" stroke="#191970" points="645.0216,-124.189 641.8458,-134.2967 650.6317,-128.3756 645.0216,-124.189"/>
</g>
<!-- Node26->Node28 -->
<g id="edge16" class="edge">
<title>Node26->Node28</title>
-<path fill="none" stroke="#191970" d="M905.9184,-142.6589C1027.6478,-129.6303 1291.0935,-101.4341 1408.9393,-88.8212"/>
-<polygon fill="#191970" stroke="#191970" points="905.2137,-139.2142 895.643,-143.7587 905.9587,-146.1745 905.2137,-139.2142"/>
+<path fill="none" stroke="#191970" d="M694.3928,-144.9071C737.8431,-141.796 796.638,-137.6102 848.5,-134 1108.0204,-115.9345 1417.7041,-95.0489 1545.198,-86.4797"/>
+<polygon fill="#191970" stroke="#191970" points="693.9744,-141.428 684.2501,-145.6338 694.4747,-148.4101 693.9744,-141.428"/>
</g>
<!-- Node29 -->
<g id="node14" class="node">
<title>Node29</title>
<g id="a_node14"><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="1786,-.5 1786,-30.5 1904,-30.5 1904,-.5 1786,-.5"/>
-<text text-anchor="start" x="1794" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/usmp</text>
-<text text-anchor="middle" x="1845" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/algo/greedy.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="1778.5,-.5 1778.5,-30.5 1896.5,-30.5 1896.5,-.5 1778.5,-.5"/>
+<text text-anchor="start" x="1786.5" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/usmp</text>
+<text text-anchor="middle" x="1837.5" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/algo/greedy.h</text>
</a>
</g>
</g>
<!-- Node28->Node29 -->
<g id="edge17" class="edge">
<title>Node28->Node29</title>
-<path fill="none" stroke="#191970" d="M1537.5475,-70.1401C1608.8119,-57.4751 1718.4326,-37.9934 1785.7143,-26.0362"/>
-<polygon fill="#191970" stroke="#191970" points="1536.5543,-66.7617 1527.321,-71.9575 1537.7791,-73.6537 1536.5543,-66.7617"/>
+<path fill="none" stroke="#191970" d="M1666.4563,-64.6842C1703.4145,-54.0568 1749.8311,-40.7095 1784.9822,-30.6017"/>
+<polygon fill="#191970" stroke="#191970" points="1665.433,-61.3366 1656.7897,-67.4639 1667.3676,-68.064 1665.433,-61.3366"/>
</g>
-<!-- Node162->Node163 -->
+<!-- Node163->Node164 -->
<g id="edge19" class="edge">
-<title>Node162->Node163</title>
-<path fill="none" stroke="#191970" d="M1104.9533,-75.1067C1229.3272,-62.1471 1485.1461,-35.491 1608.0748,-22.6819"/>
-<polygon fill="#191970" stroke="#191970" points="1104.3648,-71.649 1094.7814,-76.1666 1105.0904,-78.6113 1104.3648,-71.649"/>
+<title>Node163->Node164</title>
+<path fill="none" stroke="#191970" d="M914.5,-62.6079C914.5,-50.214 914.5,-34.8263 914.5,-25.0817"/>
+<polygon fill="#191970" stroke="#191970" points="911.0001,-62.9005 914.5,-72.9005 918.0001,-62.9006 911.0001,-62.9005"/>
</g>
-<!-- Node169->Node162 -->
+<!-- Node170->Node163 -->
<g id="edge22" class="edge">
-<title>Node169->Node162</title>
-<path fill="none" stroke="#191970" d="M1106.268,-134.0102C1088.109,-121.067 1062.7886,-103.0195 1047.4429,-92.0817"/>
-<polygon fill="#191970" stroke="#191970" points="1104.3574,-136.9464 1114.5321,-139.9005 1108.4204,-131.2462 1104.3574,-136.9464"/>
+<title>Node170->Node163</title>
+<path fill="none" stroke="#191970" d="M914.5,-129.6079C914.5,-117.214 914.5,-101.8263 914.5,-92.0817"/>
+<polygon fill="#191970" stroke="#191970" points="911.0001,-129.9005 914.5,-139.9005 918.0001,-129.9006 911.0001,-129.9005"/>
</g>
-<!-- Node171->Node135 -->
+<!-- Node172->Node135 -->
<g id="edge24" class="edge">
-<title>Node171->Node135</title>
-<path fill="none" stroke="#191970" d="M1612.2338,-320.0122C1605.5207,-300.8721 1592.5146,-272.3725 1571,-257 1542.2059,-236.4263 1449.4997,-223.2284 1387.552,-216.4682"/>
-<polygon fill="#191970" stroke="#191970" points="1608.9889,-321.3525 1615.4032,-329.785 1615.6475,-319.193 1608.9889,-321.3525"/>
+<title>Node172->Node135</title>
+<path fill="none" stroke="#191970" d="M1560.1417,-320.9346C1549.6554,-303.1923 1532.2087,-276.3571 1512.5,-257 1497.3486,-242.1189 1476.7987,-228.9773 1462.2671,-220.5915"/>
+<polygon fill="#191970" stroke="#191970" points="1557.1963,-322.8334 1565.2215,-329.7506 1563.2615,-319.3386 1557.1963,-322.8334"/>
</g>
-<!-- Node171->Node163 -->
+<!-- Node172->Node164 -->
<g id="edge28" class="edge">
-<title>Node171->Node163</title>
-<path fill="none" stroke="#191970" d="M1632.0769,-321.8208C1649.5289,-298.0732 1677,-253.8816 1677,-211 1677,-211 1677,-211 1677,-149.5 1677,-102.9722 1677,-47.357 1677,-25.2517"/>
-<polygon fill="#191970" stroke="#191970" points="1629.2197,-319.796 1625.957,-329.8759 1634.7935,-324.0307 1629.2197,-319.796"/>
+<title>Node172->Node164</title>
+<path fill="none" stroke="#191970" d="M1640.4611,-329.7871C1747.7145,-312.9074 1940.5,-273.8281 1940.5,-211 1940.5,-211 1940.5,-211 1940.5,-149.5 1940.5,-52.664 1212.6102,-23.8895 983.0121,-17.2219"/>
+<polygon fill="#191970" stroke="#191970" points="1639.7048,-326.3624 1630.3581,-331.3514 1640.7759,-333.28 1639.7048,-326.3624"/>
</g>
-<!-- Node171->Node172 -->
+<!-- Node172->Node173 -->
<g id="edge25" class="edge">
-<title>Node171->Node172</title>
-<path fill="none" stroke="#191970" d="M1592.6159,-324.711C1570.3276,-311.7256 1538.5615,-293.2184 1519.4461,-282.0817"/>
-<polygon fill="#191970" stroke="#191970" points="1591.1208,-327.8906 1601.5233,-329.9005 1594.6447,-321.8423 1591.1208,-327.8906"/>
+<title>Node172->Node173</title>
+<path fill="none" stroke="#191970" d="M1543.3395,-325.0575C1518.8902,-312.0567 1483.65,-293.3179 1462.5192,-282.0817"/>
+<polygon fill="#191970" stroke="#191970" points="1541.9747,-328.2958 1552.4473,-329.9005 1545.2612,-322.1152 1541.9747,-328.2958"/>
</g>
-<!-- Node172->Node16 -->
+<!-- Node173->Node16 -->
<g id="edge26" class="edge">
-<title>Node172->Node16</title>
-<path fill="none" stroke="#191970" d="M1496.6871,-253.0879C1487.9977,-226.3678 1472.6831,-179.2756 1466.1242,-159.1068"/>
-<polygon fill="#191970" stroke="#191970" points="1493.4402,-254.4211 1499.8613,-262.8484 1500.097,-252.2562 1493.4402,-254.4211"/>
+<title>Node173->Node16</title>
+<path fill="none" stroke="#191970" d="M1417.4296,-257.5912C1403.5854,-248.6701 1387.5743,-236.1247 1377.5,-221 1364.5412,-201.5447 1360.405,-173.614 1359.0965,-159.0922"/>
+<polygon fill="#191970" stroke="#191970" points="1415.7746,-260.6819 1426.1302,-262.921 1419.4312,-254.7128 1415.7746,-260.6819"/>
</g>
-<!-- Node172->Node135 -->
+<!-- Node173->Node135 -->
<g id="edge27" class="edge">
-<title>Node172->Node135</title>
-<path fill="none" stroke="#191970" d="M1466.0791,-259.4504C1433.0932,-247.7916 1385.5434,-230.9852 1355.9966,-220.5419"/>
-<polygon fill="#191970" stroke="#191970" points="1465.2177,-262.8581 1475.8125,-262.8906 1467.5505,-256.2582 1465.2177,-262.8581"/>
+<title>Node173->Node135</title>
+<path fill="none" stroke="#191970" d="M1444.5,-252.7462C1444.5,-242.0519 1444.5,-229.314 1444.5,-220.7449"/>
+<polygon fill="#191970" stroke="#191970" points="1441.0001,-252.8906 1444.5,-262.8906 1448.0001,-252.8907 1441.0001,-252.8906"/>
</g>
<!-- Node25->Node26 -->
<g id="edge35" class="edge">
<title>Node25->Node26</title>
-<path fill="none" stroke="#191970" d="M811.6841,-247.6276C819.3561,-222.7948 830.9003,-185.428 837.3023,-164.7056"/>
-<polygon fill="#191970" stroke="#191970" points="808.2673,-246.8302 808.6595,-257.4178 814.9554,-248.8965 808.2673,-246.8302"/>
+<path fill="none" stroke="#191970" d="M569.256,-248.6186C584.5918,-223.7987 608.1291,-185.7055 621.1047,-164.7056"/>
+<polygon fill="#191970" stroke="#191970" points="566.0981,-247.0709 563.8191,-257.4178 572.053,-250.7505 566.0981,-247.0709"/>
</g>
-<!-- Node174->Node13 -->
+<!-- Node175->Node13 -->
<g id="edge41" class="edge">
-<title>Node174->Node13</title>
-<path fill="none" stroke="#191970" d="M937.6863,-461.7211C994.6107,-448.7486 1085.0568,-428.1367 1137.8574,-416.1039"/>
-<polygon fill="#191970" stroke="#191970" points="936.7831,-458.3372 927.8108,-463.9717 938.3385,-465.1622 936.7831,-458.3372"/>
+<title>Node175->Node13</title>
+<path fill="none" stroke="#191970" d="M993.6744,-460.7484C954.0352,-447.7297 893.3995,-427.815 857.674,-416.0817"/>
+<polygon fill="#191970" stroke="#191970" points="992.6789,-464.1054 1003.2718,-463.9005 994.8632,-457.4549 992.6789,-464.1054"/>
</g>
-<!-- Node175 -->
+<!-- Node176 -->
<g id="node23" class="node">
-<title>Node175</title>
+<title>Node176</title>
<g id="a_node23"><a xlink:href="env__func_8h.html" target="_top" xlink:title="Serializable global function used in IR. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="949.5,-391.5 949.5,-421.5 1054.5,-421.5 1054.5,-391.5 949.5,-391.5"/>
-<text text-anchor="start" x="957.5" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/env</text>
-<text text-anchor="middle" x="1002" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_func.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="954,-391.5 954,-421.5 1059,-421.5 1059,-391.5 954,-391.5"/>
+<text text-anchor="start" x="962" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/env</text>
+<text text-anchor="middle" x="1006.5" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_func.h</text>
</a>
</g>
</g>
-<!-- Node174->Node175 -->
+<!-- Node175->Node176 -->
<g id="edge39" class="edge">
-<title>Node174->Node175</title>
-<path fill="none" stroke="#191970" d="M911.3929,-458.8334C930.5365,-447.7763 956.7004,-432.6644 975.9922,-421.5218"/>
-<polygon fill="#191970" stroke="#191970" points="909.5288,-455.8682 902.62,-463.9005 913.0299,-461.9297 909.5288,-455.8682"/>
+<title>Node175->Node176</title>
+<path fill="none" stroke="#191970" d="M1025.1001,-454.4311C1021.0499,-443.9939 1016.103,-431.2463 1012.3293,-421.5218"/>
+<polygon fill="#191970" stroke="#191970" points="1021.8941,-455.8441 1028.7748,-463.9005 1028.42,-453.3116 1021.8941,-455.8441"/>
</g>
-<!-- Node176 -->
+<!-- Node177 -->
<g id="node24" class="node">
-<title>Node176</title>
+<title>Node177</title>
<g id="a_node24"><a xlink:href="instrument_8h.html" target="_top" xlink:title="include/tvm/ir/instrument.h">
-<polygon fill="#ffffff" stroke="#ff0000" points="629.5,-397 629.5,-416 778.5,-416 778.5,-397 629.5,-397"/>
-<text text-anchor="middle" x="704" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/instrument.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1077,-397 1077,-416 1226,-416 1226,-397 1077,-397"/>
+<text text-anchor="middle" x="1151.5" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/ir/instrument.h</text>
</a>
</g>
</g>
-<!-- Node174->Node176 -->
+<!-- Node175->Node177 -->
<g id="edge40" class="edge">
-<title>Node174->Node176</title>
-<path fill="none" stroke="#191970" d="M850.4583,-460.416C815.0818,-447.3928 761.6304,-427.7156 730.0278,-416.0817"/>
-<polygon fill="#191970" stroke="#191970" points="849.3304,-463.7303 859.9238,-463.9005 851.7487,-457.1613 849.3304,-463.7303"/>
+<title>Node175->Node177</title>
+<path fill="none" stroke="#191970" d="M1058.4587,-458.8846C1081.5368,-445.891 1114.6133,-427.2681 1134.4818,-416.0817"/>
+<polygon fill="#191970" stroke="#191970" points="1056.5465,-455.9446 1049.5498,-463.9005 1059.9808,-462.0442 1056.5465,-455.9446"/>
</g>
-<!-- Node177 -->
+<!-- Node178 -->
<g id="node25" class="node">
-<title>Node177</title>
+<title>Node178</title>
<g id="a_node25"><a xlink:href="instruction_8h.html" target="_top" xlink:title="include/tvm/tir/schedule\l/instruction.h">
-<polygon fill="#ffffff" stroke="#ff0000" points="797,-391.5 797,-421.5 931,-421.5 931,-391.5 797,-391.5"/>
-<text text-anchor="start" x="805" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/schedule</text>
-<text text-anchor="middle" x="864" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/instruction.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="1244.5,-391.5 1244.5,-421.5 1378.5,-421.5 1378.5,-391.5 1244.5,-391.5"/>
+<text text-anchor="start" x="1252.5" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/tir/schedule</text>
+<text text-anchor="middle" x="1311.5" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/instruction.h</text>
</a>
</g>
</g>
-<!-- Node174->Node177 -->
+<!-- Node175->Node178 -->
<g id="edge42" class="edge">
-<title>Node174->Node177</title>
-<path fill="none" stroke="#191970" d="M879.6312,-454.1042C876.2243,-443.7287 872.0924,-431.145 868.9325,-421.5218"/>
-<polygon fill="#191970" stroke="#191970" points="876.4028,-455.4915 882.8479,-463.9005 883.0535,-453.3077 876.4028,-455.4915"/>
+<title>Node175->Node178</title>
+<path fill="none" stroke="#191970" d="M1082.0286,-461.606C1128.5666,-450.4302 1198.4045,-433.6591 1248.8068,-421.5553"/>
+<polygon fill="#191970" stroke="#191970" points="1081.0839,-458.2333 1072.1776,-463.9717 1082.7185,-465.0398 1081.0839,-458.2333"/>
</g>
-<!-- Node182 -->
+<!-- Node183 -->
<g id="node30" class="node">
-<title>Node182</title>
+<title>Node183</title>
<g id="a_node30"><a xlink:href="func__registry_8h.html" target="_top" xlink:title="Defines generic string-based function lookup structs. ">
-<polygon fill="#ffffff" stroke="#000000" points="306,-592.5 306,-622.5 422,-622.5 422,-592.5 306,-592.5"/>
-<text text-anchor="start" x="314" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="364" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/func_registry.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="263.5,-592.5 263.5,-622.5 379.5,-622.5 379.5,-592.5 263.5,-592.5"/>
+<text text-anchor="start" x="271.5" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="321.5" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/func_registry.h</text>
</a>
</g>
</g>
-<!-- Node181->Node182 -->
+<!-- Node182->Node183 -->
<g id="edge47" class="edge">
-<title>Node181->Node182</title>
-<path fill="none" stroke="#191970" d="M332.791,-651.0626C339.5476,-641.6315 347.1444,-631.0276 353.1512,-622.6432"/>
-<polygon fill="#191970" stroke="#191970" points="329.8706,-649.1292 326.8919,-659.2967 335.561,-653.2059 329.8706,-649.1292"/>
+<title>Node182->Node183</title>
+<path fill="none" stroke="#191970" d="M321.5,-649.0249C321.5,-640.128 321.5,-630.4287 321.5,-622.6432"/>
+<polygon fill="#191970" stroke="#191970" points="318.0001,-649.2966 321.5,-659.2967 325.0001,-649.2967 318.0001,-649.2966"/>
</g>
-<!-- Node183 -->
+<!-- Node184 -->
<g id="node31" class="node">
-<title>Node183</title>
+<title>Node184</title>
<g id="a_node31"><a xlink:href="runtime_2crt_2module_8h.html" target="_top" xlink:title="Runtime container of the functions. ">
-<polygon fill="#ffffff" stroke="#000000" points="306,-525.5 306,-555.5 422,-555.5 422,-525.5 306,-525.5"/>
-<text text-anchor="start" x="314" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="364" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/module.h</text>
+<polygon fill="#ffffff" stroke="#000000" points="177.5,-525.5 177.5,-555.5 293.5,-555.5 293.5,-525.5 177.5,-525.5"/>
+<text text-anchor="start" x="185.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="235.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/crt/module.h</text>
</a>
</g>
</g>
-<!-- Node181->Node183 -->
+<!-- Node182->Node184 -->
<g id="edge51" class="edge">
-<title>Node181->Node183</title>
-<path fill="none" stroke="#191970" d="M383.6675,-658.9598C402.1414,-651.2833 420.0614,-639.884 431,-623 438.4914,-611.4369 437.0886,-604.3595 431,-592 423.2848,-576.3386 408.265,-564.1093 394.3986,-555.5054"/>
-<polygon fill="#191970" stroke="#191970" points="382.3819,-655.7041 374.2876,-662.5403 384.8783,-662.2439 382.3819,-655.7041"/>
-</g>
-<!-- Node182->Node183 -->
-<g id="edge48" class="edge">
-<title>Node182->Node183</title>
-<path fill="none" stroke="#191970" d="M364,-582.0249C364,-573.128 364,-563.4287 364,-555.6432"/>
-<polygon fill="#191970" stroke="#191970" points="360.5001,-582.2966 364,-592.2967 367.5001,-582.2967 360.5001,-582.2966"/>
+<title>Node182->Node184</title>
+<path fill="none" stroke="#191970" d="M285.0443,-653.491C273.7239,-645.3144 262.2077,-635.0011 254.5,-623 241.1563,-602.2236 237.1436,-573.1864 235.9616,-555.6784"/>
+<polygon fill="#191970" stroke="#191970" points="283.3621,-656.58 293.5966,-659.3197 287.3044,-650.7957 283.3621,-656.58"/>
</g>
<!-- Node183->Node184 -->
-<g id="edge49" class="edge">
+<g id="edge48" class="edge">
<title>Node183->Node184</title>
-<path fill="none" stroke="#191970" d="M324.8264,-520.9132C304.2901,-510.645 279.5123,-498.2561 260.4355,-488.7177"/>
-<polygon fill="#191970" stroke="#191970" points="323.4182,-524.1222 333.9278,-525.4639 326.5488,-517.8612 323.4182,-524.1222"/>
+<path fill="none" stroke="#191970" d="M293.9855,-586.0643C281.2696,-576.1577 266.4684,-564.6265 254.9375,-555.6432"/>
+<polygon fill="#191970" stroke="#191970" points="291.9456,-588.9119 301.9853,-592.2967 296.2477,-583.3899 291.9456,-588.9119"/>
</g>
<!-- Node184->Node185 -->
-<g id="edge50" class="edge">
+<g id="edge49" class="edge">
<title>Node184->Node185</title>
-<path fill="none" stroke="#191970" d="M230,-448.0249C230,-439.128 230,-429.4287 230,-421.6432"/>
-<polygon fill="#191970" stroke="#191970" points="226.5001,-448.2966 230,-458.2967 233.5001,-448.2967 226.5001,-448.2966"/>
+<path fill="none" stroke="#191970" d="M201.2253,-520.3561C183.9439,-510.1995 163.33,-498.0843 147.3929,-488.7177"/>
+<polygon fill="#191970" stroke="#191970" points="199.5214,-523.4144 209.9162,-525.4639 203.0683,-517.3795 199.5214,-523.4144"/>
+</g>
+<!-- Node185->Node186 -->
+<g id="edge50" class="edge">
+<title>Node185->Node186</title>
+<path fill="none" stroke="#191970" d="M100.61,-450.9253C91.6871,-441.2828 81.5093,-430.2843 73.5131,-421.6432"/>
+<polygon fill="#191970" stroke="#191970" points="98.0704,-453.3342 107.4312,-458.2967 103.2082,-448.5798 98.0704,-453.3342"/>
</g>
-<!-- Node187->Node184 -->
+<!-- Node188->Node185 -->
<g id="edge56" class="edge">
-<title>Node187->Node184</title>
-<path fill="none" stroke="#191970" d="M230,-582.3415C230,-554.8131 230,-511.5714 230,-488.7614"/>
-<polygon fill="#191970" stroke="#191970" points="226.5001,-582.3889 230,-592.389 233.5001,-582.389 226.5001,-582.3889"/>
+<title>Node188->Node185</title>
+<path fill="none" stroke="#191970" d="M144.243,-582.3415C138.4908,-554.8131 129.4552,-511.5714 124.6889,-488.7614"/>
+<polygon fill="#191970" stroke="#191970" points="140.871,-583.3163 146.3425,-592.389 147.723,-581.8845 140.871,-583.3163"/>
</g>
-<!-- Node188->Node162 -->
-<g id="edge87" class="edge">
-<title>Node188->Node162</title>
-<path fill="none" stroke="#191970" d="M763.9792,-659.7518C721.3691,-649.6038 671.1996,-635.7131 653,-623 541.8143,-545.3327 522.0464,-440.1936 592,-324 653.2727,-222.2255 675.4843,-190.47 780,-134 784.5338,-131.5504 912.7261,-106.2192 985.2193,-92.0198"/>
-<polygon fill="#191970" stroke="#191970" points="763.2449,-663.1745 773.7802,-662.0527 764.8448,-656.3598 763.2449,-663.1745"/>
+<!-- Node189->Node163 -->
+<g id="edge86" class="edge">
+<title>Node189->Node163</title>
+<path fill="none" stroke="#191970" d="M743.5363,-660.1855C671.8022,-645.064 568.9473,-623.3394 568.5,-623 436.153,-522.5726 447.109,-416.8156 492.5,-257 510.0569,-195.1846 514.8384,-169.3528 568.5,-134 591.8604,-118.6099 760.8248,-98.6958 853.7022,-88.742"/>
+<polygon fill="#191970" stroke="#191970" points="742.931,-663.6348 753.4378,-662.2725 744.3747,-656.7853 742.931,-663.6348"/>
</g>
-<!-- Node188->Node169 -->
-<g id="edge88" class="edge">
-<title>Node188->Node169</title>
-<path fill="none" stroke="#191970" d="M768.8674,-655.9853C721.0966,-638.4276 658.3707,-607.0032 626,-556 570.7338,-468.9225 549.5906,-398.4146 621,-324 659.2537,-284.1365 814.4968,-305.2005 867,-288 967.9079,-254.9418 1076.2768,-184.7903 1114.0649,-159.1466"/>
-<polygon fill="#191970" stroke="#191970" points="767.9182,-659.3625 778.5129,-659.4185 770.2656,-652.7678 767.9182,-659.3625"/>
+<!-- Node189->Node170 -->
+<g id="edge87" class="edge">
+<title>Node189->Node170</title>
+<path fill="none" stroke="#191970" d="M743.5439,-657.684C713.1546,-649.058 677.3147,-637.3286 646.5,-623 596.3034,-599.6591 568.1088,-604.5434 541.5,-556 534.8774,-543.9182 538.5934,-538.4677 541.5,-525 562.1878,-429.1448 567.773,-397.6655 632.5,-324 707.744,-238.3651 835.2889,-180.8098 889.2204,-159.1258"/>
+<polygon fill="#191970" stroke="#191970" points="742.6334,-661.0635 753.2058,-660.3749 744.5115,-654.3202 742.6334,-661.0635"/>
</g>
-<!-- Node188->Node172 -->
+<!-- Node189->Node173 -->
<g id="edge58" class="edge">
-<title>Node188->Node172</title>
-<path fill="none" stroke="#191970" d="M876.1981,-655.4628C897.7059,-646.0073 923.8539,-634.2275 947,-623 1004.773,-594.976 1021.4048,-591.364 1075,-556 1093.1925,-543.996 1094.6357,-536.7395 1113,-525 1258.3697,-432.0712 1339.5164,-476.5148 1462,-355 1483.4598,-333.71 1495.5803,-298.8886 1500.4547,-282.1201"/>
-<polygon fill="#191970" stroke="#191970" points="874.7365,-652.2819 866.9795,-659.4986 877.5437,-658.6944 874.7365,-652.2819"/>
+<title>Node189->Node173</title>
+<path fill="none" stroke="#191970" d="M845.9244,-653.6155C858.1209,-645.1293 871.327,-634.5864 881.5,-623 903.9437,-597.438 907.3684,-588.1438 918.5,-556 942.8167,-485.7826 890.9013,-441.4059 945.5,-391 1016.9319,-325.0535 1296.2061,-401.652 1381.5,-355 1411.8571,-338.396 1432.1161,-300.2917 1440.3482,-282.2774"/>
+<polygon fill="#191970" stroke="#191970" points="843.6331,-650.938 837.2798,-659.4167 847.5337,-656.7506 843.6331,-650.938"/>
</g>
-<!-- Node188->Node174 -->
+<!-- Node189->Node175 -->
<g id="edge59" class="edge">
-<title>Node188->Node174</title>
-<path fill="none" stroke="#191970" d="M791.212,-654.7546C739.2745,-626.9491 660.5059,-574.1464 698,-525 712.9329,-505.4262 771.6214,-491.4662 819.5039,-483.0616"/>
-<polygon fill="#191970" stroke="#191970" points="789.7637,-657.947 800.2461,-659.4865 793.0117,-651.7461 789.7637,-657.947"/>
+<title>Node189->Node175</title>
+<path fill="none" stroke="#191970" d="M768.6074,-654.1149C756.6716,-646.1807 745.1624,-635.8536 738.5,-623 727.5166,-601.8098 742.3426,-532.2822 749.5,-525 777.242,-496.7744 879.1154,-483.7583 952.6903,-477.9379"/>
+<polygon fill="#191970" stroke="#191970" points="766.9598,-657.2106 777.3107,-659.4711 770.6287,-651.2491 766.9598,-657.2106"/>
</g>
-<!-- Node189 -->
+<!-- Node190 -->
<g id="node37" class="node">
-<title>Node189</title>
+<title>Node190</title>
<g id="a_node37"><a xlink:href="structural__equal_8h.html" target="_top" xlink:title="Structural equality comparison. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="706.5,-525.5 706.5,-555.5 857.5,-555.5 857.5,-525.5 706.5,-525.5"/>
-<text text-anchor="start" x="714.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/structural</text>
-<text text-anchor="middle" x="782" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_equal.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="550,-525.5 550,-555.5 701,-555.5 701,-525.5 550,-525.5"/>
+<text text-anchor="start" x="558" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/structural</text>
+<text text-anchor="middle" x="625.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_equal.h</text>
</a>
</g>
</g>
-<!-- Node188->Node189 -->
+<!-- Node189->Node190 -->
<g id="edge60" class="edge">
-<title>Node188->Node189</title>
-<path fill="none" stroke="#191970" d="M805.1398,-652.0203C797.1809,-643.7847 789.4184,-633.809 785,-623 775.9977,-600.9772 777.4088,-572.8263 779.5137,-555.7547"/>
-<polygon fill="#191970" stroke="#191970" points="802.7965,-654.6234 812.4138,-659.0684 807.6676,-649.5961 802.7965,-654.6234"/>
+<title>Node189->Node190</title>
+<path fill="none" stroke="#191970" d="M743.5704,-658.181C720.5389,-650.2802 695.7859,-638.9426 676.5,-623 654.3069,-604.6542 639.0071,-574.0072 631.2937,-555.6782"/>
+<polygon fill="#191970" stroke="#191970" points="742.7222,-661.5858 753.3146,-661.3601 744.8933,-654.931 742.7222,-661.5858"/>
</g>
-<!-- Node190 -->
+<!-- Node191 -->
<g id="node38" class="node">
-<title>Node190</title>
+<title>Node191</title>
<g id="a_node38"><a xlink:href="structural__hash_8h.html" target="_top" xlink:title="include/tvm/node/structural\l_hash.h">
-<polygon fill="#ffffff" stroke="#ff0000" points="914.5,-525.5 914.5,-555.5 1065.5,-555.5 1065.5,-525.5 914.5,-525.5"/>
-<text text-anchor="start" x="922.5" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/structural</text>
-<text text-anchor="middle" x="990" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_hash.h</text>
+<polygon fill="#ffffff" stroke="#ff0000" points="758,-525.5 758,-555.5 909,-555.5 909,-525.5 758,-525.5"/>
+<text text-anchor="start" x="766" y="-543.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/node/structural</text>
+<text text-anchor="middle" x="833.5" y="-532.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_hash.h</text>
</a>
</g>
</g>
-<!-- Node188->Node190 -->
+<!-- Node189->Node191 -->
<g id="edge63" class="edge">
-<title>Node188->Node190</title>
-<path fill="none" stroke="#191970" d="M871.8432,-654.5227C887.317,-645.8646 904.7159,-634.9663 919,-623 943.4804,-602.4919 966.5813,-573.1069 979.4553,-555.5028"/>
-<polygon fill="#191970" stroke="#191970" points="870.0942,-651.4899 862.9931,-659.3528 873.4476,-657.6344 870.0942,-651.4899"/>
+<title>Node189->Node191</title>
+<path fill="none" stroke="#191970" d="M774.5533,-653.6709C763.8098,-645.597 753.3902,-635.3002 747.5,-623 741.5493,-610.5735 740.7066,-603.9865 747.5,-592 756.8532,-575.497 773.8324,-563.7109 790.2481,-555.6168"/>
+<polygon fill="#191970" stroke="#191970" points="772.5606,-656.5483 782.7623,-659.4081 776.5706,-650.8107 772.5606,-656.5483"/>
</g>
-<!-- Node188->Node191 -->
+<!-- Node189->Node192 -->
<g id="edge66" class="edge">
-<title>Node188->Node191</title>
-<path fill="none" stroke="#191970" d="M900.0238,-658.7878C967.314,-643.7986 1033.83,-629.5453 1079.9926,-619.7434"/>
-<polygon fill="#191970" stroke="#191970" points="899.09,-655.41 890.0924,-661.0042 900.6147,-662.2419 899.09,-655.41"/>
+<title>Node189->Node192</title>
+<path fill="none" stroke="#191970" d="M877.7349,-656.8198C917.6302,-646.1704 967.8972,-632.7525 1005.925,-622.6017"/>
+<polygon fill="#191970" stroke="#191970" points="876.5883,-653.5032 867.8293,-659.4639 878.3937,-660.2664 876.5883,-653.5032"/>
</g>
-<!-- Node188->Node196 -->
-<g id="edge85" class="edge">
-<title>Node188->Node196</title>
-<path fill="none" stroke="#191970" d="M899.9707,-658.9722C1035.764,-636.502 1075.4264,-656.0754 1205,-623 1240.7454,-613.8755 1246.2425,-601.077 1282,-592 1318.9456,-582.6214 1558.9495,-557.7077 1671.6802,-546.3229"/>
-<polygon fill="#191970" stroke="#191970" points="899.2963,-655.5366 890.0249,-660.6641 900.4703,-662.4375 899.2963,-655.5366"/>
+<!-- Node189->Node197 -->
+<g id="edge84" class="edge">
+<title>Node189->Node197</title>
+<path fill="none" stroke="#191970" d="M879.6813,-664.2823C965.873,-651.2375 1105.9197,-629.5679 1129.5,-623 1165.0387,-613.1012 1170.7543,-601.1233 1206.5,-592 1276.9779,-574.0121 1484.1091,-554.3566 1586.2266,-545.4173"/>
+<polygon fill="#191970" stroke="#191970" points="879.0718,-660.8346 869.7071,-665.7898 880.1179,-667.756 879.0718,-660.8346"/>
</g>
-<!-- Node204 -->
-<g id="node47" class="node">
-<title>Node204</title>
-<g id="a_node47"><a xlink:href="bytecode_8h.html" target="_top" xlink:title="The bytecode for Relay virtual machine. ">
-<polygon fill="#ffffff" stroke="#ff0000" points="794,-592.5 794,-622.5 910,-622.5 910,-592.5 794,-592.5"/>
-<text text-anchor="start" x="802" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="852" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/vm/bytecode.h</text>
+<!-- Node205 -->
+<g id="node46" class="node">
+<title>Node205</title>
+<g id="a_node46"><a xlink:href="bytecode_8h.html" target="_top" xlink:title="The bytecode for Relay virtual machine. ">
+<polygon fill="#ffffff" stroke="#ff0000" points="756.5,-592.5 756.5,-622.5 872.5,-622.5 872.5,-592.5 756.5,-592.5"/>
+<text text-anchor="start" x="764.5" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
+<text text-anchor="middle" x="814.5" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/vm/bytecode.h</text>
</a>
</g>
</g>
-<!-- Node188->Node204 -->
-<g id="edge86" class="edge">
-<title>Node188->Node204</title>
-<path fill="none" stroke="#191970" d="M839.4298,-649.6103C842.1328,-640.5553 845.1045,-630.5998 847.4797,-622.6432"/>
-<polygon fill="#191970" stroke="#191970" points="836.045,-648.7133 836.5383,-659.2967 842.7525,-650.7156 836.045,-648.7133"/>
+<!-- Node189->Node205 -->
+<g id="edge85" class="edge">
+<title>Node189->Node205</title>
+<path fill="none" stroke="#191970" d="M812.6407,-649.0249C813.039,-640.128 813.4733,-630.4287 813.8219,-622.6432"/>
+<polygon fill="#191970" stroke="#191970" points="809.1317,-649.1501 812.1807,-659.2967 816.1246,-649.4633 809.1317,-649.1501"/>
</g>
-<!-- Node189->Node13 -->
+<!-- Node190->Node13 -->
<g id="edge61" class="edge">
-<title>Node189->Node13</title>
-<path fill="none" stroke="#191970" d="M856.0206,-522.9975C892.2032,-513.8733 936.2058,-501.9286 975,-489 1043.425,-466.1966 1121.9087,-432.3929 1158.7752,-416.0436"/>
-<polygon fill="#191970" stroke="#191970" points="854.8857,-519.6735 846.0334,-525.4947 856.5838,-526.4644 854.8857,-519.6735"/>
+<title>Node190->Node13</title>
+<path fill="none" stroke="#191970" d="M656.8058,-519.8351C701.3538,-490.429 781.2549,-437.6864 813.9504,-416.1042"/>
+<polygon fill="#191970" stroke="#191970" points="654.8096,-516.9589 648.3921,-525.389 658.666,-522.8009 654.8096,-516.9589"/>
</g>
-<!-- Node189->Node174 -->
+<!-- Node190->Node175 -->
<g id="edge62" class="edge">
-<title>Node189->Node174</title>
-<path fill="none" stroke="#191970" d="M813.7519,-520.0445C832.8201,-507.7601 856.1932,-492.7025 871.0924,-483.1039"/>
-<polygon fill="#191970" stroke="#191970" points="811.8506,-517.1058 805.3396,-525.4639 815.6417,-522.9904 811.8506,-517.1058"/>
+<title>Node190->Node175</title>
+<path fill="none" stroke="#191970" d="M711.0649,-525.7439C712.5546,-525.4935 714.034,-525.2453 715.5,-525 805.1364,-510.0005 909.1417,-493.2232 972.8784,-483.0131"/>
+<polygon fill="#191970" stroke="#191970" points="710.3909,-522.3081 701.1132,-527.4241 711.5564,-529.2104 710.3909,-522.3081"/>
</g>
-<!-- Node190->Node13 -->
+<!-- Node191->Node13 -->
<g id="edge64" class="edge">
-<title>Node190->Node13</title>
-<path fill="none" stroke="#191970" d="M1019.8239,-519.4663C1061.6186,-489.99 1135.908,-437.5965 1166.3822,-416.1042"/>
-<polygon fill="#191970" stroke="#191970" points="1017.5809,-516.7653 1011.4261,-525.389 1021.6154,-522.4857 1017.5809,-516.7653"/>
+<title>Node191->Node13</title>
+<path fill="none" stroke="#191970" d="M832.5581,-515.2559C831.4378,-485.2327 829.6234,-436.6073 828.8584,-416.1042"/>
+<polygon fill="#191970" stroke="#191970" points="829.0656,-515.5265 832.9362,-525.389 836.0608,-515.2654 829.0656,-515.5265"/>
</g>
-<!-- Node190->Node174 -->
+<!-- Node191->Node175 -->
<g id="edge65" class="edge">
-<title>Node190->Node174</title>
-<path fill="none" stroke="#191970" d="M958.2481,-520.0445C939.1799,-507.7601 915.8068,-492.7025 900.9076,-483.1039"/>
-<polygon fill="#191970" stroke="#191970" points="956.3583,-522.9904 966.6604,-525.4639 960.1494,-517.1058 956.3583,-522.9904"/>
+<title>Node191->Node175</title>
+<path fill="none" stroke="#191970" d="M887.9658,-522.1623C925.5961,-509.4928 974.1179,-493.1563 1004.2679,-483.0053"/>
+<polygon fill="#191970" stroke="#191970" points="886.5199,-518.9559 878.1595,-525.4639 888.7536,-525.59 886.5199,-518.9559"/>
</g>
-<!-- Node191->Node8 -->
+<!-- Node192->Node8 -->
<g id="edge67" class="edge">
-<title>Node191->Node8</title>
-<path fill="none" stroke="#191970" d="M1069.6762,-602.88C953.813,-594.5494 728.3468,-576.0074 698,-556 610.1224,-498.0631 602,-444.7576 602,-339.5 602,-339.5 602,-339.5 602,-211 602,-169.8905 605.6162,-121.4926 607.6503,-97.5206"/>
-<polygon fill="#191970" stroke="#191970" points="1069.6353,-606.3858 1079.859,-603.6061 1070.1333,-599.4036 1069.6353,-606.3858"/>
+<title>Node192->Node8</title>
+<path fill="none" stroke="#191970" d="M1101.752,-587.9093C1184.6468,-546.1461 1371.303,-449.8943 1387.5,-422 1394.4184,-410.0852 1394.3426,-402.9585 1387.5,-391 1358.3126,-339.991 1328.1512,-345.612 1273.5,-324 974.8042,-205.8798 858.8906,-302.2766 568.5,-165 532.2378,-147.8578 496.8528,-116.5086 477.5357,-97.7813"/>
+<polygon fill="#191970" stroke="#191970" points="1100.0077,-584.8688 1092.6459,-592.4881 1103.1524,-591.1227 1100.0077,-584.8688"/>
</g>
-<!-- Node191->Node174 -->
+<!-- Node192->Node175 -->
<g id="edge68" class="edge">
-<title>Node191->Node174</title>
-<path fill="none" stroke="#191970" d="M1195.3379,-588.709C1232.2037,-573.6327 1269.3651,-550.7821 1247,-525 1228.8247,-504.0478 1066.3562,-487.6473 965.6697,-479.3856"/>
-<polygon fill="#191970" stroke="#191970" points="1194.0375,-585.4594 1186.008,-592.3716 1196.5954,-591.9753 1194.0375,-585.4594"/>
+<title>Node192->Node175</title>
+<path fill="none" stroke="#191970" d="M1100.8523,-586.6919C1111.8317,-578.6618 1122.4545,-568.3855 1128.5,-556 1134.5436,-543.6185 1135.5742,-536.823 1128.5,-525 1115.7784,-503.7387 1090.9383,-490.7515 1069.7136,-483.1377"/>
+<polygon fill="#191970" stroke="#191970" points="1098.7636,-583.8795 1092.4566,-592.3925 1102.6958,-589.6707 1098.7636,-583.8795"/>
</g>
-<!-- Node191->Node190 -->
+<!-- Node192->Node191 -->
<g id="edge69" class="edge">
-<title>Node191->Node190</title>
-<path fill="none" stroke="#191970" d="M1095.6173,-588.3132C1072.6739,-577.9267 1044.7366,-565.2794 1023.359,-555.6017"/>
-<polygon fill="#191970" stroke="#191970" points="1094.2325,-591.5282 1104.7859,-592.4639 1097.1194,-585.1512 1094.2325,-591.5282"/>
+<title>Node192->Node191</title>
+<path fill="none" stroke="#191970" d="M1001.1431,-589.5484C964.8955,-578.9432 919.5169,-565.6665 885.1162,-555.6017"/>
+<polygon fill="#191970" stroke="#191970" points="1000.5274,-593.015 1011.1079,-592.4639 1002.4931,-586.2966 1000.5274,-593.015"/>
</g>
-<!-- Node191->Node193 -->
+<!-- Node192->Node194 -->
<g id="edge70" class="edge">
-<title>Node191->Node193</title>
-<path fill="none" stroke="#191970" d="M1206.326,-596.0839C1361.2751,-570.1947 1737.0906,-507.4026 1881.831,-483.219"/>
-<polygon fill="#191970" stroke="#191970" points="1205.2869,-592.7089 1196.0005,-597.8091 1206.4406,-599.6132 1205.2869,-592.7089"/>
+<title>Node192->Node194</title>
+<path fill="none" stroke="#191970" d="M1130.7461,-591.1752C1160.9262,-582.6754 1196.3339,-570.909 1226.5,-556 1247.9455,-545.401 1248.2434,-533.7686 1270.5,-525 1326.7551,-502.8367 1496.1842,-485.8699 1586.438,-478.1213"/>
+<polygon fill="#191970" stroke="#191970" points="1129.4396,-587.9046 1120.7266,-593.9324 1131.2969,-594.6537 1129.4396,-587.9046"/>
</g>
-<!-- Node191->Node196 -->
+<!-- Node192->Node197 -->
<g id="edge74" class="edge">
-<title>Node191->Node196</title>
-<path fill="none" stroke="#191970" d="M1206.1297,-597.5368C1219.9749,-595.6143 1234.4467,-593.6785 1248,-592 1400.6307,-573.0979 1580.9755,-554.9382 1671.6846,-546.0936"/>
-<polygon fill="#191970" stroke="#191970" points="1205.5989,-594.0769 1196.1811,-598.9304 1206.57,-601.0092 1205.5989,-594.0769"/>
+<title>Node192->Node197</title>
+<path fill="none" stroke="#191970" d="M1130.6327,-597.5606C1144.4776,-595.6363 1158.9486,-593.6936 1172.5,-592 1321.3659,-573.395 1497.1271,-555.2142 1586.3002,-546.258"/>
+<polygon fill="#191970" stroke="#191970" points="1130.1016,-594.1007 1120.6841,-598.9547 1131.0731,-601.033 1130.1016,-594.1007"/>
</g>
-<!-- Node191->Node202 -->
-<g id="edge82" class="edge">
-<title>Node191->Node202</title>
-<path fill="none" stroke="#191970" d="M1146.5514,-583.4837C1151.8767,-574.1996 1158.6104,-563.8565 1164.7537,-555.6432"/>
-<polygon fill="#191970" stroke="#191970" points="1143.4624,-581.8379 1141.7696,-592.2967 1149.6151,-585.1763 1143.4624,-581.8379"/>
+<!-- Node192->Node203 -->
+<g id="edge81" class="edge">
+<title>Node192->Node203</title>
+<path fill="none" stroke="#191970" d="M1055.5626,-582.3179C1055.0141,-573.3414 1055.1368,-563.5143 1055.9725,-555.6432"/>
+<polygon fill="#191970" stroke="#191970" points="1052.0805,-582.6733 1056.5123,-592.2967 1059.049,-582.01 1052.0805,-582.6733"/>
</g>
-<!-- Node191->Node203 -->
-<g id="edge84" class="edge">
-<title>Node191->Node203</title>
-<path fill="none" stroke="#191970" d="M1206.1631,-599.6341C1230.1199,-597.0098 1257.2358,-594.1976 1282,-592 1510.5493,-571.7186 1568.7291,-579.2059 1797,-556 1812.3624,-554.4383 1828.827,-552.4377 1844.3854,-550.3966"/>
-<polygon fill="#191970" stroke="#191970" points="1205.6607,-596.1682 1196.1054,-600.7449 1206.4292,-603.1259 1205.6607,-596.1682"/>
+<!-- Node192->Node204 -->
+<g id="edge83" class="edge">
+<title>Node192->Node204</title>
+<path fill="none" stroke="#191970" d="M1130.4915,-591.516C1176.9342,-580.5979 1238.068,-566.2261 1283.6105,-555.5196"/>
+<polygon fill="#191970" stroke="#191970" points="1129.4511,-588.1651 1120.5175,-593.8608 1131.0531,-594.9793 1129.4511,-588.1651"/>
</g>
-<!-- Node193->Node28 -->
+<!-- Node194->Node28 -->
<g id="edge73" class="edge">
-<title>Node193->Node28</title>
-<path fill="none" stroke="#191970" d="M1904.2561,-452.8841C1891.2907,-444.3586 1877.1136,-433.7196 1866,-422 1786.4807,-338.1445 1815.8032,-277.2907 1729,-201 1668.9131,-148.1901 1581.3757,-114.9779 1524.5538,-97.5743"/>
-<polygon fill="#191970" stroke="#191970" points="1902.6696,-456.023 1912.986,-458.436 1906.4261,-450.1162 1902.6696,-456.023"/>
+<title>Node194->Node28</title>
+<path fill="none" stroke="#191970" d="M1648.731,-448.1914C1652.8522,-421.3097 1658.5,-377.5324 1658.5,-339.5 1658.5,-339.5 1658.5,-339.5 1658.5,-211 1658.5,-167.2001 1631.5058,-121.0997 1615.6888,-97.8472"/>
+<polygon fill="#191970" stroke="#191970" points="1645.2405,-447.8567 1647.1333,-458.2811 1652.1543,-448.9516 1645.2405,-447.8567"/>
</g>
-<!-- Node193->Node29 -->
+<!-- Node194->Node29 -->
<g id="edge72" class="edge">
-<title>Node193->Node29</title>
-<path fill="none" stroke="#191970" d="M1922.2251,-450.0754C1904.5338,-424.3808 1880,-381.2261 1880,-339.5 1880,-339.5 1880,-339.5 1880,-149.5 1880,-105.2584 1861.9189,-55.3895 1851.7484,-30.8378"/>
-<polygon fill="#191970" stroke="#191970" points="1919.4994,-452.282 1928.1459,-458.4048 1925.2048,-448.2263 1919.4994,-452.282"/>
+<title>Node194->Node29</title>
+<path fill="none" stroke="#191970" d="M1676.332,-452.9069C1749.3452,-404.8865 1924.803,-284.739 1953.5,-221 1969.3751,-185.7397 1969.1842,-169.7726 1954.5,-134 1935.6168,-87.998 1889.2922,-50.3184 1860.9248,-30.544"/>
+<polygon fill="#191970" stroke="#191970" points="1674.399,-449.989 1667.9525,-458.3969 1678.2352,-455.8442 1674.399,-449.989"/>
</g>
-<!-- Node193->Node194 -->
+<!-- Node194->Node195 -->
<g id="edge71" class="edge">
-<title>Node193->Node194</title>
-<path fill="none" stroke="#191970" d="M1883.9677,-455.2759C1851.5015,-444.7165 1811.1131,-431.5805 1780.4321,-421.6017"/>
-<polygon fill="#191970" stroke="#191970" points="1883.1774,-458.6993 1893.7696,-458.4639 1885.3425,-452.0425 1883.1774,-458.6993"/>
+<title>Node194->Node195</title>
+<path fill="none" stroke="#191970" d="M1701.1928,-455.4123C1734.3618,-444.8298 1775.7547,-431.6235 1807.1664,-421.6017"/>
+<polygon fill="#191970" stroke="#191970" points="1700.0911,-452.0899 1691.6281,-458.4639 1702.2188,-458.7587 1700.0911,-452.0899"/>
</g>
-<!-- Node196->Node6 -->
+<!-- Node197->Node6 -->
<g id="edge75" class="edge">
-<title>Node196->Node6</title>
-<path fill="none" stroke="#191970" d="M1661.8294,-538.2543C1543.0951,-532.0758 1292.8881,-508.9027 1103,-422 1082.8662,-412.7857 1084.372,-399.6749 1064,-391 953.4266,-343.9153 914.7946,-373.2019 796,-355 791.4551,-354.3036 786.7849,-353.5661 782.0763,-352.8057"/>
-<polygon fill="#191970" stroke="#191970" points="1661.7103,-541.7526 1671.8714,-538.7524 1662.0571,-534.7612 1661.7103,-541.7526"/>
+<title>Node197->Node6</title>
+<path fill="none" stroke="#191970" d="M1616.7406,-519.2317C1571.4811,-485.7845 1477.755,-421.5596 1387.5,-391 1266.9883,-350.1957 1229.2415,-374.4195 1103.5,-355 1100.6833,-354.565 1097.8166,-354.1092 1094.9226,-353.638"/>
+<polygon fill="#191970" stroke="#191970" points="1614.765,-522.1245 1624.8735,-525.2978 1618.9502,-516.5134 1614.765,-522.1245"/>
</g>
-<!-- Node196->Node174 -->
+<!-- Node197->Node175 -->
<g id="edge76" class="edge">
-<title>Node196->Node174</title>
-<path fill="none" stroke="#191970" d="M1661.486,-535.3149C1541.439,-526.1957 1285.2655,-506.5835 1069,-489 1035.1546,-486.2482 997.798,-483.1034 965.7092,-480.3689"/>
-<polygon fill="#191970" stroke="#191970" points="1661.395,-538.818 1671.6313,-536.0852 1661.9251,-531.8381 1661.395,-538.818"/>
+<title>Node197->Node175</title>
+<path fill="none" stroke="#191970" d="M1576.3701,-533.0413C1462.363,-520.5602 1233.5226,-495.5074 1112.2177,-482.2273"/>
+<polygon fill="#191970" stroke="#191970" points="1576.0512,-536.5273 1586.3728,-534.1364 1576.8131,-529.5688 1576.0512,-536.5273"/>
</g>
-<!-- Node196->Node193 -->
+<!-- Node197->Node194 -->
<g id="edge77" class="edge">
-<title>Node196->Node193</title>
-<path fill="none" stroke="#191970" d="M1786.6928,-522.4123C1819.8618,-511.8298 1861.2547,-498.6235 1892.6664,-488.6017"/>
-<polygon fill="#191970" stroke="#191970" points="1785.5911,-519.0899 1777.1281,-525.4639 1787.7188,-525.7587 1785.5911,-519.0899"/>
-</g>
-<!-- Node196->Node194 -->
-<g id="edge81" class="edge">
-<title>Node196->Node194</title>
-<path fill="none" stroke="#191970" d="M1730.751,-515.3415C1731.5727,-487.8131 1732.8635,-444.5714 1733.5444,-421.7614"/>
-<polygon fill="#191970" stroke="#191970" points="1727.2511,-515.289 1730.4511,-525.389 1734.248,-515.4979 1727.2511,-515.289"/>
-</g>
-<!-- Node196->Node197 -->
-<g id="edge78" class="edge">
-<title>Node196->Node197</title>
-<path fill="none" stroke="#191970" d="M1797.9729,-527.0084C1852.1934,-516.9076 1929.9184,-502.5545 1998,-489 2005.6241,-487.4821 2013.6843,-485.8243 2021.6735,-484.1755"/>
-<polygon fill="#191970" stroke="#191970" points="1797.3189,-523.57 1788.13,-528.8438 1798.6021,-530.4514 1797.3189,-523.57"/>
+<title>Node197->Node194</title>
+<path fill="none" stroke="#191970" d="M1644.5,-515.0249C1644.5,-506.128 1644.5,-496.4287 1644.5,-488.6432"/>
+<polygon fill="#191970" stroke="#191970" points="1641.0001,-515.2966 1644.5,-525.2967 1648.0001,-515.2967 1641.0001,-515.2966"/>
</g>
-<!-- Node197->Node196 -->
+<!-- Node197->Node195 -->
<g id="edge80" class="edge">
-<title>Node197->Node196</title>
-<path fill="none" stroke="#191970" d="M2011.8744,-489.8058C1930.3552,-505.9344 1843.7837,-521.7337 1788.0273,-531.3499"/>
-<polygon fill="#191970" stroke="#191970" points="2012.6613,-493.218 2021.7883,-487.8377 2011.2981,-486.352 2012.6613,-493.218"/>
-</g>
-<!-- Node198 -->
-<g id="node44" class="node">
-<title>Node198</title>
-<g id="a_node44"><a xlink:href="libtorch__runtime_8h.html" target="_top" xlink:title="include/tvm/runtime\l/contrib/libtorch_runtime.h">
-<polygon fill="#ffffff" stroke="#000000" points="2007,-391.5 2007,-421.5 2153,-421.5 2153,-391.5 2007,-391.5"/>
-<text text-anchor="start" x="2015" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">include/tvm/runtime</text>
-<text text-anchor="middle" x="2080" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">/contrib/libtorch_runtime.h</text>
-</a>
-</g>
+<title>Node197->Node195</title>
+<path fill="none" stroke="#191970" d="M1672.5112,-519.256C1684.5705,-510.0361 1698.7911,-499.0648 1711.5,-489 1728.5866,-475.4682 1731.1517,-469.7645 1749.5,-458 1771.5489,-443.8628 1798.0384,-430.8963 1818.8945,-421.5104"/>
+<polygon fill="#191970" stroke="#191970" points="1670.381,-516.4788 1664.5528,-525.3265 1674.6264,-522.0444 1670.381,-516.4788"/>
</g>
<!-- Node197->Node198 -->
-<g id="edge79" class="edge">
+<g id="edge78" class="edge">
<title>Node197->Node198</title>
-<path fill="none" stroke="#191970" d="M2080,-448.0249C2080,-439.128 2080,-429.4287 2080,-421.6432"/>
-<polygon fill="#191970" stroke="#191970" points="2076.5001,-448.2966 2080,-458.2967 2083.5001,-448.2967 2076.5001,-448.2966"/>
+<path fill="none" stroke="#191970" d="M1686.9005,-521.5897C1712.7745,-511.1518 1745.7314,-498.3654 1771.983,-488.6017"/>
+<polygon fill="#191970" stroke="#191970" points="1685.3081,-518.4587 1677.3596,-525.4639 1687.9417,-524.9444 1685.3081,-518.4587"/>
</g>
-<!-- Node202->Node191 -->
-<g id="edge83" class="edge">
-<title>Node202->Node191</title>
-<path fill="none" stroke="#191970" d="M1171.4876,-564.4483C1166.1692,-573.7272 1159.4374,-584.0729 1153.2913,-592.2967"/>
-<polygon fill="#191970" stroke="#191970" points="1174.572,-566.1026 1176.2608,-555.6432 1168.4181,-562.7665 1174.572,-566.1026"/>
+<!-- Node198->Node197 -->
+<g id="edge79" class="edge">
+<title>Node198->Node197</title>
+<path fill="none" stroke="#191970" d="M1773.9255,-492.4805C1748.0198,-502.9289 1715.0621,-515.7137 1688.8407,-525.4639"/>
+<polygon fill="#191970" stroke="#191970" points="1775.531,-495.6062 1783.48,-488.6017 1772.8979,-489.1203 1775.531,-495.6062"/>
</g>
-<!-- Node206->Node13 -->
-<g id="edge98" class="edge">
-<title>Node206->Node13</title>
-<path fill="none" stroke="#191970" d="M1544.609,-652.3186C1492.4738,-603.205 1367.4471,-487.3064 1318,-458 1285.5518,-438.7685 1244.6827,-424.6017 1215.8684,-416.0476"/>
-<polygon fill="#191970" stroke="#191970" points="1542.3669,-655.0152 1552.0417,-659.3336 1547.1715,-649.9244 1542.3669,-655.0152"/>
+<!-- Node203->Node192 -->
+<g id="edge82" class="edge">
+<title>Node203->Node192</title>
+<path fill="none" stroke="#191970" d="M1068.4332,-565.6135C1068.9851,-574.5857 1068.866,-584.4156 1068.0339,-592.2967"/>
+<polygon fill="#191970" stroke="#191970" points="1071.9159,-565.2645 1067.4796,-555.6432 1064.9477,-565.931 1071.9159,-565.2645"/>
</g>
-<!-- Node206->Node16 -->
-<g id="edge94" class="edge">
-<title>Node206->Node16</title>
-<path fill="none" stroke="#191970" d="M1583.5132,-650.6038C1588.5897,-642.116 1593.9573,-632.3455 1598,-623 1640.7102,-524.2671 1611.9467,-483.4202 1667,-391 1678.58,-371.5602 1694.3409,-375.905 1703,-355 1708.2725,-342.271 1709.2528,-336.2772 1703,-324 1657.2927,-234.2551 1541.4294,-179.814 1489.2187,-159.0945"/>
-<polygon fill="#191970" stroke="#191970" points="1580.4639,-648.8806 1578.1954,-659.2298 1586.4226,-652.5541 1580.4639,-648.8806"/>
+<!-- Node206->Node160 -->
+<g id="edge91" class="edge">
+<title>Node206->Node160</title>
+<path fill="none" stroke="#191970" d="M1934.1212,-650.4837C1939.3868,-641.1996 1945.253,-630.8565 1949.9113,-622.6432"/>
+<polygon fill="#191970" stroke="#191970" points="1931.0118,-648.8716 1929.1228,-659.2967 1937.1007,-652.325 1931.0118,-648.8716"/>
</g>
-<!-- Node206->Node135 -->
+<!-- Node208->Node13 -->
+<g id="edge99" class="edge">
+<title>Node208->Node13</title>
+<path fill="none" stroke="#191970" d="M1530.9423,-650.4484C1508.1114,-610.3059 1463.3225,-533.2156 1451.5,-525 1405.955,-493.35 1046.1336,-438.1084 894.9927,-416.022"/>
+<polygon fill="#191970" stroke="#191970" points="1528.0045,-652.3633 1535.9816,-659.336 1534.0938,-648.9106 1528.0045,-652.3633"/>
+</g>
+<!-- Node208->Node16 -->
<g id="edge95" class="edge">
-<title>Node206->Node135</title>
-<path fill="none" stroke="#191970" d="M1556.6309,-649.9769C1545.4352,-623.5463 1530,-579.9532 1530,-540.5 1530,-540.5 1530,-540.5 1530,-406.5 1530,-354.2911 1388.8858,-252.32 1343.1339,-220.6311"/>
-<polygon fill="#191970" stroke="#191970" points="1553.4601,-651.4617 1560.6732,-659.222 1559.8738,-648.6573 1553.4601,-651.4617"/>
+<title>Node208->Node16</title>
+<path fill="none" stroke="#191970" d="M1612.7734,-667.3129C1743.5414,-649.3714 2016.5,-589.9498 2016.5,-406.5 2016.5,-406.5 2016.5,-406.5 2016.5,-339.5 2016.5,-214.6516 1573.7235,-166.7792 1414.6007,-153.6061"/>
+<polygon fill="#191970" stroke="#191970" points="1612.3014,-663.8447 1602.8499,-668.6321 1613.2239,-670.7837 1612.3014,-663.8447"/>
</g>
-<!-- Node206->Node171 -->
+<!-- Node208->Node135 -->
<g id="edge96" class="edge">
-<title>Node206->Node171</title>
-<path fill="none" stroke="#191970" d="M1568,-649.1092C1568,-622.1594 1568,-578.3265 1568,-540.5 1568,-540.5 1568,-540.5 1568,-473.5 1568,-424.087 1598.2132,-370.5245 1611.6675,-349.1221"/>
-<polygon fill="#191970" stroke="#191970" points="1564.5001,-649.227 1568,-659.2271 1571.5001,-649.2271 1564.5001,-649.227"/>
+<title>Node208->Node135</title>
+<path fill="none" stroke="#191970" d="M1612.5509,-657.2148C1644.076,-648.3667 1681.7268,-636.5902 1714.5,-623 1743.6126,-610.9278 1814.7235,-575.4784 1839.5,-556 1903.8776,-505.3888 1968.0746,-458.3557 1921.5,-391 1869.7745,-316.1951 1593.7222,-245.4983 1486.8238,-220.5171"/>
+<polygon fill="#191970" stroke="#191970" points="1611.2493,-653.9434 1602.5447,-659.9833 1613.116,-660.6899 1611.2493,-653.9434"/>
</g>
-<!-- Node206->Node172 -->
+<!-- Node208->Node172 -->
<g id="edge97" class="edge">
-<title>Node206->Node172</title>
-<path fill="none" stroke="#191970" d="M1575.1103,-649.3193C1577.3981,-640.9662 1579.8861,-631.6042 1582,-623 1607.1881,-520.4757 1573.4247,-478.1577 1633,-391 1649.277,-367.187 1673.3895,-380.4314 1687,-355 1693.5012,-342.8525 1695.3265,-334.9771 1687,-324 1671.8141,-303.9801 1610.161,-289.6191 1562.3574,-281.2306"/>
-<polygon fill="#191970" stroke="#191970" points="1571.6623,-648.6557 1572.3628,-659.2274 1578.4078,-650.5263 1571.6623,-648.6557"/>
+<title>Node208->Node172</title>
+<path fill="none" stroke="#191970" d="M1546.4868,-648.9003C1551.8182,-580.2081 1566.2704,-393.9966 1569.7425,-349.26"/>
+<polygon fill="#191970" stroke="#191970" points="1542.9691,-648.9944 1545.6847,-659.2352 1549.9481,-649.5361 1542.9691,-648.9944"/>
</g>
-<!-- Node206->Node174 -->
-<g id="edge99" class="edge">
-<title>Node206->Node174</title>
-<path fill="none" stroke="#191970" d="M1499.668,-669.8148C1383.0389,-661.418 1150.4544,-642.8221 1071,-623 994.2052,-603.8414 957.679,-615.9481 906,-556 887.8128,-534.9027 885.579,-499.9895 885.6554,-483.1597"/>
-<polygon fill="#191970" stroke="#191970" points="1499.6794,-673.3245 1509.9037,-670.547 1500.1789,-666.3423 1499.6794,-673.3245"/>
+<!-- Node208->Node173 -->
+<g id="edge98" class="edge">
+<title>Node208->Node173</title>
+<path fill="none" stroke="#191970" d="M1538.5755,-649.4341C1524.914,-591.8417 1489.959,-445.6692 1458.5,-324 1454.7425,-309.4679 1450.1343,-292.7047 1447.2567,-282.3536"/>
+<polygon fill="#191970" stroke="#191970" points="1535.2339,-650.5115 1540.945,-659.4353 1542.0453,-648.8978 1535.2339,-650.5115"/>
</g>
-<!-- Node206->Node179 -->
+<!-- Node208->Node175 -->
<g id="edge100" class="edge">
-<title>Node206->Node179</title>
-<path fill="none" stroke="#191970" d="M1515.2621,-656.0003C1479.2102,-643.3538 1432.9119,-627.1131 1404.0972,-617.0053"/>
-<polygon fill="#191970" stroke="#191970" points="1514.541,-659.4564 1525.1359,-659.4639 1516.8582,-652.851 1514.541,-659.4564"/>
+<title>Node208->Node175</title>
+<path fill="none" stroke="#191970" d="M1512.3133,-653.5507C1500.7386,-645 1488.1665,-634.4306 1478.5,-623 1446.5255,-585.1902 1466.2936,-553.0684 1425.5,-525 1376.1092,-491.0163 1211.8772,-479.4354 1112.0235,-475.5045"/>
+<polygon fill="#191970" stroke="#191970" points="1510.3375,-656.4406 1520.5095,-659.4037 1514.4055,-650.7439 1510.3375,-656.4406"/>
</g>
-<!-- Node206->Node191 -->
-<g id="edge102" class="edge">
-<title>Node206->Node191</title>
-<path fill="none" stroke="#191970" d="M1499.6298,-663.847C1415.8224,-650.7886 1275.2409,-628.8841 1196.3115,-616.5857"/>
-<polygon fill="#191970" stroke="#191970" points="1499.1411,-667.313 1509.5608,-665.3944 1500.2189,-660.3964 1499.1411,-667.313"/>
+<!-- Node208->Node180 -->
+<g id="edge101" class="edge">
+<title>Node208->Node180</title>
+<path fill="none" stroke="#191970" d="M1480.0356,-656.7259C1433.8108,-643.9807 1373.339,-627.3075 1335.9744,-617.0053"/>
+<polygon fill="#191970" stroke="#191970" points="1479.3954,-660.1799 1489.9661,-659.4639 1481.2561,-653.4317 1479.3954,-660.1799"/>
</g>
-<!-- Node206->Node194 -->
+<!-- Node208->Node192 -->
<g id="edge104" class="edge">
-<title>Node206->Node194</title>
-<path fill="none" stroke="#191970" d="M1635.8453,-660.6274C1690.6898,-645.3528 1764.0094,-614.7227 1797,-556 1823.2111,-509.3446 1775.8121,-449.5924 1749.6297,-421.8837"/>
-<polygon fill="#191970" stroke="#191970" points="1634.9336,-657.2481 1626.179,-663.2154 1636.744,-664.01 1634.9336,-657.2481"/>
+<title>Node208->Node192</title>
+<path fill="none" stroke="#191970" d="M1476.272,-665.016C1381.1617,-651.7953 1210.3573,-628.0528 1120.8467,-615.6104"/>
+<polygon fill="#191970" stroke="#191970" points="1475.9729,-668.508 1486.3595,-666.4182 1476.9367,-661.5747 1475.9729,-668.508"/>
+</g>
+<!-- Node208->Node195 -->
+<g id="edge106" class="edge">
+<title>Node208->Node195</title>
+<path fill="none" stroke="#191970" d="M1592.2879,-655.645C1677.9023,-620.7661 1851.5115,-544.6583 1883.5,-489 1896.1338,-467.0179 1879.9434,-438.6075 1867.1591,-421.5103"/>
+<polygon fill="#191970" stroke="#191970" points="1590.749,-652.492 1582.7946,-659.4904 1593.3771,-658.98 1590.749,-652.492"/>
+</g>
+<!-- Node208->Node197 -->
+<g id="edge105" class="edge">
+<title>Node208->Node197</title>
+<path fill="none" stroke="#191970" d="M1561.929,-651.1452C1582.3832,-623.7365 1615.7119,-579.076 1633.1109,-555.7614"/>
+<polygon fill="#191970" stroke="#191970" points="1558.9527,-649.2813 1555.7769,-659.389 1564.5628,-653.4679 1558.9527,-649.2813"/>
</g>
-<!-- Node206->Node196 -->
+<!-- Node208->Node198 -->
<g id="edge103" class="edge">
-<title>Node206->Node196</title>
-<path fill="none" stroke="#191970" d="M1609.2111,-654.6743C1625.2176,-646.0372 1643.2175,-635.1138 1658,-623 1682.8497,-602.6364 1706.2681,-573.2051 1719.3153,-555.5529"/>
-<polygon fill="#191970" stroke="#191970" points="1607.2791,-651.7357 1600.0565,-659.4872 1610.5365,-657.9317 1607.2791,-651.7357"/>
+<title>Node208->Node198</title>
+<path fill="none" stroke="#191970" d="M1574.845,-653.6311C1608.5555,-630.2946 1664.4379,-591.1532 1711.5,-556 1742.1462,-533.1087 1776.9067,-505.4535 1797.7468,-488.6849"/>
+<polygon fill="#191970" stroke="#191970" points="1572.5976,-650.9297 1566.361,-659.4944 1576.5775,-656.6882 1572.5976,-650.9297"/>
</g>
-<!-- Node206->Node197 -->
-<g id="edge101" class="edge">
-<title>Node206->Node197</title>
-<path fill="none" stroke="#191970" d="M1636.3436,-668.7332C1697.326,-662.1602 1788.401,-648.798 1864,-623 1924.3978,-602.3894 1938.9301,-592.8974 1991,-556 2019.8798,-535.5354 2049.1839,-506.4008 2065.918,-488.789"/>
-<polygon fill="#191970" stroke="#191970" points="1635.6528,-665.2863 1626.0709,-669.8069 1636.3806,-672.2483 1635.6528,-665.2863"/>
+<!-- Node208->Node204 -->
+<g id="edge107" class="edge">
+<title>Node208->Node204</title>
+<path fill="none" stroke="#191970" d="M1498.8416,-655.1081C1480.1327,-646.3527 1458.6559,-635.228 1440.5,-623 1410.2462,-602.6241 1379.4069,-573.1968 1361.9399,-555.5486"/>
+<polygon fill="#191970" stroke="#191970" points="1497.5093,-658.3476 1508.0578,-659.3368 1500.4285,-651.9853 1497.5093,-658.3476"/>
</g>
-<!-- Node206->Node203 -->
-<g id="edge105" class="edge">
-<title>Node206->Node203</title>
-<path fill="none" stroke="#191970" d="M1636.1571,-665.6749C1702.2103,-656.3112 1797.0946,-640.4768 1830,-623 1861.042,-606.5129 1888.0332,-574.633 1902.2723,-555.694"/>
-<polygon fill="#191970" stroke="#191970" points="1635.6431,-662.2127 1626.2248,-667.0652 1636.6135,-669.1451 1635.6431,-662.2127"/>
+<!-- Node208->Node160 -->
+<g id="edge102" class="edge">
+<title>Node208->Node160</title>
+<path fill="none" stroke="#191970" d="M1612.4949,-663.496C1692.6624,-650.522 1824.5413,-629.1793 1900.2244,-616.9311"/>
+<polygon fill="#191970" stroke="#191970" points="1611.8132,-660.0607 1602.5008,-665.1134 1612.9316,-666.9708 1611.8132,-660.0607"/>
</g>
</g>
</svg>
diff --git a/docs/reference/api/doxygen/classes.html b/docs/reference/api/doxygen/classes.html
index 1a875fcf8..50f04824e 100644
--- a/docs/reference/api/doxygen/classes.html
+++ b/docs/reference/api/doxygen/classes.html
@@ -65,227 +65,228 @@ $(function() {
<div class="qindex"><a class="qindex" href="#letter_a">a</a> | <a class="qindex" href="#letter_b">b</a> | <a class="qindex" href="#letter_c">c</a> | <a class="qindex" href="#letter_d">d</a> | <a class="qindex" href="#letter_e">e</a> | <a class="qindex" href="#letter_f">f</a> | <a class="qindex" href="#letter_g">g</a> | <a class="qindex" href="#letter_h">h</a> | <a class="qindex" href="#letter_i">i</a> |& [...]
<table class="classindex">
<tr><td rowspan="2" valign="bottom"><a name="letter_a"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  a  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DWinogradAttrs.html">Conv3DWinogradAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1IterAdapter.html">IterAdapter</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1PrefetchNode.html">PrefetchNode</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConvGemmWeightTransformAttrs.html">ConvGemmWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1MapNode_1_1iterator.html">MapNode::iterator</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1Preload [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzer.html">AccessAnalyzer</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConvWinogradWeightTransformAttrs.html">ConvWinogradWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1run [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzerNode.html">AccessAnalyzerNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CorrelationAttrs.html">CorrelationAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1Iterator [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool1DAttrs.html">AdaptivePool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1CostModel.html">CostModel</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1support_1_1Span_1_1iterator__base.html">Sp [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool2DAttrs.html">AdaptivePool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModel.html">CostModel</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1IteratorNode.html">I [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool3DAttrs.html">AdaptivePool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1CostModelNode.html">CostModelNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1auto__scheduler_1_1AttachMapNode_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Add.html">Add</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModelNode.html">CostModelNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMapExpr.html">IterMapExpr</a> (<a class="el" href="namesp [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AddNode.html">AddNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1CountNode.html">CountNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMapExprNode.html">IterMapExprNode</a> (<a c [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ADT.html">ADT</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CropAndResizeAttrs.html">CropAndResizeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMark.html">IterMark</a> (<a class="el" href="namespacetvm_1_1ar [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ADTObj.html">ADTObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_d"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  d  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMarkNode.html">IterMarkNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ProducerLoad.html">ProducerLoad</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1StepNode.html">StepNode</a> (<a class="el" href="namespacetvm_1 [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AffineGridAttrs.html">AffineGridAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSplitExpr.html">IterSplitExpr</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ProducerLoadNode.html">ProducerLoadNode</a> (<a class="el" hr [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AffineType.html">AffineType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Database.html">Database</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSplitExprNode.html">IterSplitExprNode</a> (<a class="el" href="namespacetvm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AffineTypeNode.html">AffineTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1DatabaseNode.html">DatabaseNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSumExpr.html">IterSumExpr</a> (<a class="el" href="namespac [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllClassNonMaximumSuppressionAttrs.html">AllClassNonMaximumSuppressionAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DataProducer.html">DataProducer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSumExprNode.html">IterSum [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Allocate.html">Allocate</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DataProducerNode.html">DataProducerNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IterVar.html">IterVar</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir< [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateConst.html">AllocateConst</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DataType.html">DataType</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarAttr.html">IterVarAttr</a> (<a class="el" href="namespacetvm_1_1te.htm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateConstNode.html">AllocateConstNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DataTypePattern.html">DataTypePattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarAttrNode.html">IterVarAttrNode</a> (<a class="el" href [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1AllocatedPoolInfo.html">AllocatedPoolInfo</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DataTypePatternNode.html">DataTypePatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IterVarNode.html">IterVarNode< [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1AllocatedPoolInfoNode.html">AllocatedPoolInfoNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DebugAttrs.html">DebugAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarRelation.html">IterVarRelation</ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateNode.html">AllocateNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DeformableConv2DAttrs.html">DeformableConv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarRelationNode.html">IterVarRelationNode</a> (<a clas [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1Allocator.html">Allocator</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DenseAttrs.html">DenseAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_l"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td> [...]
-</td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramRunner.html">ProgramRunner</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html">StmtVisitor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllocStorageAttrs.html">AllocStorageAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DenseMapNode.html">DenseMapNode</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramRunnerNode.html">ProgramRunnerNode [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllocTensorAttrs.html">AllocTensorAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DensePackAttrs.html">DensePackAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1L2NormalizeAttrs.html">L2NormalizeAttrs</a> (<a class [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AltPattern.html">AltPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Dependency.html">Dependency</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LayerNormAttrs.html">LayerNormAttrs</a> (<a class="el" href="namespacetvm_1_1rela [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AltPatternNode.html">AltPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DependencyNode.html">DependencyNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Layout.html">Layout</a> (<a class="el" href="namespacetvm_1_1tir.htm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1Analyzer.html">Analyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1qnn_1_1DequantizeAttrs.html">DequantizeAttrs</a> (<a class="el" href="namespacetvm_1_1relay_1_1qnn.html">tvm::relay::qnn</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LayoutAxis.html">LayoutAxis</a> (<a class="el" href= [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1And.html">And</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DeviceAPI.html">DeviceAPI</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LayoutNode.html">LayoutNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)&# [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AndNode.html">AndNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DeviceCopyAttrs.html">DeviceCopyAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LayoutTransformAttrs.html">LayoutTransformAttrs</a> (<a class="el" href="nam [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStep.html">AnnotationStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1DeviceWrapper.html">DeviceWrapper</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStepNode.html">AnnotationStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1profiling_1_1DeviceWrapperNode.html">DeviceWrapperNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Any.html">Any</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPattern.html">DFPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1tir_1_1LENode.html">LENode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AnyNode.html">AnyNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternCallback.html">DFPatternCallback</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Let.html">Let</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::re [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ApplyHistoryBest.html">ApplyHistoryBest</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternCallbackNode.html">DFPatternCallbackNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Let.html">Let</a> [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ApplyHistoryBestNode.html">ApplyHistoryBestNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor.html">DFPatternFunctor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1LetNode.html">LetN [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArangeAttrs.html">ArangeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor_3_01R_07const_01DFPattern_01_6n_00_01Args_8_8_8_08_4.html">DFPatternFunctor< R(const DFPattern &n, Args...)></a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a cla [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfo.html">ArgInfo</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternNode.html">DFPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1LetPattern.html">LetPattern</a> (<a class="el" hre [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfoNode.html">ArgInfoNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternVisitor.html">DFPatternVisitor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1LetPatternNode.html">LetPatternNode< [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArgReduceAttrs.html">ArgReduceAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Diagnostic.html">Diagnostic</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LetStmt.html">LetStmt</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>) &# [...]
+</td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConvGemmWeightTransformAttrs.html">ConvGemmWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Map_1_1iterator.html">Map::iterator</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1PreloadMeasure [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConvWinogradWeightTransformAttrs.html">ConvWinogradWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1Iterator.html">Iterator</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzer.html">AccessAnalyzer</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CorrelationAttrs.html">CorrelationAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1support_1_1Span_1_1iterator__base.h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzerNode.html">AccessAnalyzerNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1CostModel.html">CostModel</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_ [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool1DAttrs.html">AdaptivePool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModel.html">CostModel</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1auto__scheduler_1_1AttachMapNode_1_1It [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool2DAttrs.html">AdaptivePool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1CostModelNode.html">CostModelNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMapExpr.html">IterMap [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AdaptivePool3DAttrs.html">AdaptivePool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModelNode.html">CostModelNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMapExprNode.html"> [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Add.html">Add</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1CountNode.html">CountNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMark.html">IterMark</a> (<a class="el" href="namesp [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AddNode.html">AddNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CropAndResizeAttrs.html">CropAndResizeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterMarkNode.html">IterMarkNode</a> (<a class="el" href="namespacetvm_1 [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ADT.html">ADT</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_d"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  d  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSplitExpr.html">IterSplitExpr</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ProducerLoad.html">ProducerLoad</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1StftAttrs.html">StftAttrs</a> (<a class="el" href="namespacetvm_1_1rel [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ADTObj.html">ADTObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSplitExprNode.html">IterSplitExprNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ProducerLoadNode.html">ProducerLoadNode</a> (<a class="el" href="n [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AffineGridAttrs.html">AffineGridAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Database.html">Database</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSumExpr.html">IterSumExpr</a> (<a class [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AffineType.html">AffineType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1DatabaseNode.html">DatabaseNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IterSumExprNode.html">IterSumExprNode</a> (<a class="el" href="namespac [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AffineTypeNode.html">AffineTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DataProducer.html">DataProducer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IterVar.html">IterVar</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllClassNonMaximumSuppressionAttrs.html">AllClassNonMaximumSuppressionAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DataProducerNode.html">DataProducerNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarAttr.html">IterVa [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Allocate.html">Allocate</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DataType.html">DataType</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarAttrNode.html">IterVarAttrNode</a> (<a class="el" href="namespacetvm_1_1te.html" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateConst.html">AllocateConst</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DataTypePattern.html">DataTypePattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IterVarNode.html">IterVarNode</a> (<a class="el" href="namespacetvm_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateConstNode.html">AllocateConstNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DataTypePatternNode.html">DataTypePatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarRelation.html">IterVarRelation</a> (<a class=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1AllocatedPoolInfo.html">AllocatedPoolInfo</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DebugAttrs.html">DebugAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1IterVarRelationNode.html">IterVarRelationNode</a [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1AllocatedPoolInfoNode.html">AllocatedPoolInfoNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DeformableConv2DAttrs.html">DeformableConv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_l"></a><table border="0" [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramMeasurerNode.html">ProgramMeasurerNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html">StmtVisitor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AllocateNode.html">AllocateNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DenseAttrs.html">DenseAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramRunner.html">ProgramRunner</a> (<a class="el" href="namespac [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1Allocator.html">Allocator</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DenseMapNode.html">DenseMapNode</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1L2NormalizeAttrs.html">L2NormalizeAttrs</a> ( [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllocStorageAttrs.html">AllocStorageAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DensePackAttrs.html">DensePackAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LayerNormAttrs.html">LayerNormAttrs</a> (<a class=" [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AllocTensorAttrs.html">AllocTensorAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Dependency.html">Dependency</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Layout.html">Layout</a> (<a class="el" href="namespacetvm_1_1tir.html"> [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AltPattern.html">AltPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DependencyNode.html">DependencyNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LayoutAxis.html">LayoutAxis</a> (<a class="el" href="namespacetvm_1_1tir.htm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AltPatternNode.html">AltPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1qnn_1_1DequantizeAttrs.html">DequantizeAttrs</a> (<a class="el" href="namespacetvm_1_1relay_1_1qnn.html">tvm::relay::qnn</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LayoutNode.html">LayoutNode</a> (<a clas [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1Analyzer.html">Analyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1DeviceAPI.html">DeviceAPI</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LayoutTransformAttrs.html">LayoutTransformAttrs</a> (<a class="el" href="na [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1And.html">And</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DeviceCopyAttrs.html">DeviceCopyAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LE.html">LE</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AndNode.html">AndNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1DeviceWrapper.html">DeviceWrapper</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LeakyReluAttrs.html">LeakyReluAttrs</a [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStep.html">AnnotationStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1profiling_1_1DeviceWrapperNode.html">DeviceWrapperNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="st [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStepNode.html">AnnotationStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPattern.html">DFPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Let.html">Let</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Any.html">Any</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternCallback.html">DFPatternCallback</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Let.html">Let</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)  [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AnyNode.html">AnyNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternCallbackNode.html">DFPatternCallbackNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1LetNode.html">LetNode</a> (<a class="el" href="namespacetvm_1_1rel [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ApplyHistoryBest.html">ApplyHistoryBest</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor.html">DFPatternFunctor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LetNode.html">LetNode</a> (< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ApplyHistoryBestNode.html">ApplyHistoryBestNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor_3_01R_07const_01DFPattern_01_6n_00_01Args_8_8_8_08_4.html">DFPatternFunctor< R(const DFPattern &n, Args...)></a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>) [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArangeAttrs.html">ArangeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternNode.html">DFPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1LetPatternNode.html">LetPatternNode</a> (<a class="el" href="namesp [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfo.html">ArgInfo</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DFPatternVisitor.html">DFPatternVisitor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LetStmt.html">LetStmt</a> (<a class="el" href= [...]
</td></tr>
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArgsortAttrs.html">ArgsortAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticBuilder.html">DiagnosticBuilder</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LetStmtNode.html">LetStmtNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Array.html">Array</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticContext.html">DiagnosticContext</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1support_1_1LinearCongruentialEngine.html">LinearCongruentialEngine</a> (<a class="el" href="namespac [...]
-</td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1TakeAttrs.html">TakeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html">ArrayAccessor</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticContextNode.html">DiagnosticContextNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Load.html">Load</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html">ArrayAccessor< const char *, ::tvm::runtime::String ></a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticNode.html">DiagnosticNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </t [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1SimpleObjAllocator_1_1ArrayHandler.html">SimpleObjAllocator::ArrayHandler</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticRenderer.html">DiagnosticRenderer</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalBuilder.html">Loca [...]
-</td><td valign="top"><a class="el" href="classtvm_1_1TargetKindAttrMap.html">TargetKindAttrMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html">ArrayIterator</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticRendererNode.html">DiagnosticRendererNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalBuilderNode.html [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ArrayNode.html">ArrayNode</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DictAttrs.html">DictAttrs</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunner.html">LocalRunner</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.ht [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AssertStmt.html">AssertStmt</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DictAttrsNode.html">DictAttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunnerNode.html">LocalRunnerNode</a> (<a class="el" href="namespacetvm_1_1auto__schedu [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AssertStmtNode.html">AssertStmtNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DilateAttrs.html">DilateAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LoopRV.html">LoopRV</a> (<a class="el" href="namespacetvm_1_1tir.html">tv [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMap.html">AttachMap</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Dilation2DAttrs.html">Dilation2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LoopRVNode.html">LoopRVNode</a> (<a clas [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMapNode.html">AttachMapNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Div.html">Div</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LRNAttrs.html">LRNAttrs</a> (<a class="el" href="namespacetv [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrDocEntry.html">AttrDocEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DivNode.html">DivNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LT.html">LT</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>) & [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrDocVisitor.html">AttrDocVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DominatorPattern.html">DominatorPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LTNode.html">LTNode</a> (<a class="el" href="namespacet [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1AttrError.html">AttrError</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DominatorPatternNode.html">DominatorPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_m"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">   [...]
-</td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Rebase.html">Rebase</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1TempExpr.html">TempExpr</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrExistVisitor.html">AttrExistVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DropoutAttrs.html">DropoutAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1RebaseNode.html">RebaseNode</a> (<a class="el" href="namesp [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrFieldInfo.html">AttrFieldInfo</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1DurationNode.html">DurationNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a> (<a class="el" href="namespa [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrFieldInfoNode.html">AttrFieldInfoNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DynExpandDimsAttrs.html">DynExpandDimsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1MapNode.html">MapNode</a> (<a class="el" href="namespacetvm_1_1runti [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrInitEntry.html">AttrInitEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_e"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  e  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Match.html">Match</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReader.html">RecordReader</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TensorAffineTypeNode.html">TensorAffineTypeNode</a> (<a class [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrInitVisitor.html">AttrInitVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MatchBufferRegion.html">MatchBufferRegion</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReaderNode.html">RecordReaderNode</a> ( [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrNonDefaultVisitor.html">AttrNonDefaultVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1EinsumAttrs.html">EinsumAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MatchBufferRegionNode.html">MatchBufferRegionNode< [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrNopEntry.html">AttrNopEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1EnvFunc.html">EnvFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MatchNode.html">MatchNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)  [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrNormalVisitor.html">AttrNormalVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1EnvFuncNode.html">EnvFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MatmulAttrs.html">MatmulAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AttrPattern.html">AttrPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EQ.html">EQ</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MatrixSetDiagAttrs.html">MatrixSetDiagAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AttrPatternNode.html">AttrPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EQNode.html">EQNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Max.html">Max</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>) [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistry.html">AttrRegistry</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1ErrorBuilder.html">ErrorBuilder</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MaxNode.html">MaxNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistryMap.html">AttrRegistryMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1ErrorReporter.html">ErrorReporter</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool1DAttrs.html">MaxPool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)  [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistryMapContainerMap.html">AttrRegistryMapContainerMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Evaluate.html">Evaluate</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool2DAttrs.html">MaxPool2DAttrs</a> (<a class="el" href="namespacetvm_1_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1Attrs.html">Attrs</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EvaluateNode.html">EvaluateNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool3DAttrs.html">MaxPool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>) &# [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrsNode.html">AttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1Executable.html">Executable</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" href="namespace [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrsSEqualVisitor.html">AttrsSEqualVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Executor.html">Executor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallback.html">MeasureCallback</a> (<a class [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrsSHashVisitor.html">AttrsSHashVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExecutorNode.html">ExecutorNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallbackNode.html">MeasureCallbackNode< [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AttrStmt.html">AttrStmt</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExecutorRegEntry.html">ExecutorRegEntry</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallbackNode.html">MeasureCallbackNode</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AttrStmtNode.html">AttrStmtNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ExpandDimsAttrs.html">ExpandDimsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidate.html">MeasureCandidate</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrTriggerNonDefaultEntry.html">AttrTriggerNonDefaultEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1tir_1_1ExprDeepEqual.html">ExprDeepEqual</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidateNode.html">Mea [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttrVisitor.html">AttrVisitor</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprFunctor.html">ExprFunctor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInput.html">MeasureInput</a> (<a class="el" href="namespacetvm_1_1auto__schedule [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AutoSchedulerLayoutTransformAttrs.html">AutoSchedulerLayoutTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprFunctor.html">ExprFunctor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInputNode.html"> [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool1DAttrs.html">AvgPool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html">ExprFunctor< R(const Expr &n, Args...)></a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href=" [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool2DAttrs.html">AvgPool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html">ExprFunctor< R(const PrimExpr &n, Args...)></a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool3DAttrs.html">AvgPool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprMutator.html">ExprMutator</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1MemoryInfo.html">MemoryInfo</a> (<a class="el" href="namespacetvm.html">tv [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfoNode.html">ArgInfoNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Diagnostic.html">Diagnostic</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LetStmtNode.html">LetStmtNode</a> (<a class="el" href="namespacetvm_1_1tir. [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArgReduceAttrs.html">ArgReduceAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticBuilder.html">DiagnosticBuilder</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1support_1_1LinearCongruentialEngine.html">LinearCongruentialEngine</a> (<a class="el" h [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ArgsortAttrs.html">ArgsortAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticContext.html">DiagnosticContext</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Load.html">Load</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>) [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1Target.html">Target</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Array.html">Array</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticContextNode.html">DiagnosticContextNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LoadNode.html">LoadNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html">ArrayAccessor</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticNode.html">DiagnosticNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalBuilder.html">LocalBuilder</a> ( [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html">ArrayAccessor< const char *, ::tvm::runtime::String ></a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticRenderer.html">DiagnosticRenderer</a> (<a class="el" href="namespacetvm.html">tvm</a>)  & [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1TargetKindNode.html">TargetKindNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1SimpleObjAllocator_1_1ArrayHandler.html">SimpleObjAllocator::ArrayHandler</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DiagnosticRendererNode.html">DiagnosticRendererNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunner.htm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html">ArrayIterator</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm::runtime::metadata</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DictAttrs.html">DictAttrs</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunnerNode.html">LocalRunnerNode</a> (<a c [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ArrayNode.html">ArrayNode</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1DictAttrsNode.html">DictAttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LoopRV.html">LoopRV</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)  [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AssertStmt.html">AssertStmt</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DilateAttrs.html">DilateAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LoopRVNode.html">LoopRVNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tv [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AssertStmtNode.html">AssertStmtNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Dilation2DAttrs.html">Dilation2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1LRNAttrs.html">LRNAttrs</a> (<a class="el" href="namespacetvm_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMap.html">AttachMap</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Div.html">Div</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LT.html">LT</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMapNode.html">AttachMapNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1DivNode.html">DivNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1LTNode.html">LTNode</a> (<a class="el" href="namespacet [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrDocEntry.html">AttrDocEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DominatorPattern.html">DominatorPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_m"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><d [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1RatioNode.html">RatioNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1TempExpr.html">TempExpr</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrDocVisitor.html">AttrDocVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1DominatorPatternNode.html">DominatorPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Rebase.html">Rebase</a> (<a class="el" href="nam [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1AttrError.html">AttrError</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DropoutAttrs.html">DropoutAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)  [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrExistVisitor.html">AttrExistVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1DurationNode.html">DurationNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1MapNode.html [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrFieldInfo.html">AttrFieldInfo</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1DynExpandDimsAttrs.html">DynExpandDimsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Match.html">Match</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::re [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrFieldInfoNode.html">AttrFieldInfoNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_e"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  e  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MatchBufferRegion.html">MatchBufferRegion</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReader.html">RecordReader</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1TensorComputeOp.html">TensorComputeOp [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrInitEntry.html">AttrInitEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MatchBufferRegionNode.html">MatchBufferRegionNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReaderNode.html">RecordReaderNode< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrInitVisitor.html">AttrInitVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1EinsumAttrs.html">EinsumAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MatchNode.html">MatchNode</a> (<a class="el" href="namespace [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrNonDefaultVisitor.html">AttrNonDefaultVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1EnvFunc.html">EnvFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MatmulAttrs.html">MatmulAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrNopEntry.html">AttrNopEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1EnvFuncNode.html">EnvFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MatrixSetDiagAttrs.html">MatrixSetDiagAttrs</a> (<a class="el" href="namespacetvm_1_1relay [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrNormalVisitor.html">AttrNormalVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EQ.html">EQ</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Max.html">Max</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)  [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AttrPattern.html">AttrPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EQNode.html">EQNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MaxNode.html">MaxNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>) [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1AttrPatternNode.html">AttrPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1ErrorBuilder.html">ErrorBuilder</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool1DAttrs.html">MaxPool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.ht [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistry.html">AttrRegistry</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1ErrorReporter.html">ErrorReporter</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool2DAttrs.html">MaxPool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistryMap.html">AttrRegistryMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Evaluate.html">Evaluate</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool3DAttrs.html">MaxPool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrRegistryMapContainerMap.html">AttrRegistryMapContainerMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1EvaluateNode.html">EvaluateNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" hre [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1Attrs.html">Attrs</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1Executable.html">Executable</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" href="namespacetvm_1_1 [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrsNode.html">AttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Executor.html">Executor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallbackNode.html">MeasureCallbackNode</a> (<a class="el" href="namespacetvm_1_1meta__sched [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrsSEqualVisitor.html">AttrsSEqualVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExecutorNode.html">ExecutorNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallbackNode.html">MeasureCallbackNo [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1detail_1_1AttrsSHashVisitor.html">AttrsSHashVisitor</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExecutorRegEntry.html">ExecutorRegEntry</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidate.html">MeasureCandidat [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AttrStmt.html">AttrStmt</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ExpandDimsAttrs.html">ExpandDimsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidateNode.html">MeasureCandidateNode</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1AttrStmtNode.html">AttrStmtNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1tir_1_1ExprDeepEqual.html">ExprDeepEqual</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInput.html">MeasureInput</a> (<a class="el" href="namespacet [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1detail_1_1AttrTriggerNonDefaultEntry.html">AttrTriggerNonDefaultEntry</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprFunctor.html">ExprFunctor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInputNode.html">Measu [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttrVisitor.html">AttrVisitor</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprFunctor.html">ExprFunctor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureResult.html">MeasureResult</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.ht [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AutoSchedulerLayoutTransformAttrs.html">AutoSchedulerLayoutTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html">ExprFunctor< R(const Expr &n, Args...)></a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool1DAttrs.html">AvgPool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html">ExprFunctor< R(const PrimExpr &n, Args...)></a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool2DAttrs.html">AvgPool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprMutator.html">ExprMutator</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1MemoryInfoNode.html">MemoryInfoNode</a> (<a class="el" href="namespacetvm. [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1AvgPool3DAttrs.html">AvgPool3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprMutator.html">ExprMutator</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1MemoryManager.html">MemoryManager</a> (<a class="el" href="name [...]
<tr><td rowspan="2" valign="bottom"><a name="letter_b"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  b  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprMutator.html">ExprMutator</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1MemoryInfoNode.html">MemoryInfoNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1RelayRefType.html">RelayRefType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </t [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprPattern.html">ExprPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1vm_1_1MemoryManager.html">MemoryManager</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1RelayRefTypeNode.html">RelayRefTypeNode</a> (<a class=" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseAttrsNode.html">BaseAttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprPatternNode.html">ExprPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structMemoryManagerInterface.html">MemoryManagerInterface</a>   </td><td valign="top"><a cl [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1BaseComputeOpNode.html">BaseComputeOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprRewriter.html">ExprRewriter</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MeshgridAttrs.html">MeshgridAttrs</a> (<a class="el" href="namespa [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseExpr.html">BaseExpr</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html">ExprVisitor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html">MetadataArray</a> (<a class="el" href="namespacetvm_1_1runtime_1_1met [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseExprNode.html">BaseExprNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html">ExprVisitor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html">MetadataArrayNode</a> (<a class="el" href="namespacetvm_1_1runt [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseFunc.html">BaseFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1ExternOp.html">ExternOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">tvm:: [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseFuncNode.html">BaseFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1ExternOpNode.html">ExternOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html">MetadataBaseNode</a> (<a class="el" href="namespacetvm_1_1runtime [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseTensorType.html">BaseTensorType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTask.html">ExtractedTask</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollector.html">MetricCollector</a> (<a [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseTensorTypeNode.html">BaseTensorTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTaskNode.html">ExtractedTaskNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollectorNode.html">Metr [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseValueEqual.html">BaseValueEqual</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncObj_1_1Extractor.html">PackedFuncObj::Extractor</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Min.html">Min</a> (<a class="el" href="namespacetvm_1_1tir [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1BaseValueHash.html">BaseValueHash</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_f"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  f  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MinNode.html">MinNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ReshapeTensorAttrs.html">ReshapeTensorAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TupleAffineType.html">TupleAffineType</a> (<a class="el" href="namespacetvm.htm [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchMatmulAttrs.html">BatchMatmulAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MirrorPadAttrs.html">MirrorPadAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Resize1DAttrs.html">Resize1DAttrs</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchNormAttrs.html">BatchNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractor.html">FeatureExtractor</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html">MixedM [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchToSpaceNDAttrs.html">BatchToSpaceNDAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractorNode.html">FeatureExtractorNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MixedModeVi [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BiasAddAttrs.html">BiasAddAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FeatureSet.html">FeatureSet</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Mod.html">Mod</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BijectiveLayout.html">BijectiveLayout</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1FIFOBufferAttrs.html">FIFOBufferAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ModNode.html">ModNode</a> (<a class="el" href="namespacetvm_1_1 [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BijectiveLayoutNode.html">BijectiveLayoutNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1FixedPointMultiplyAttrs.html">FixedPointMultiplyAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSet.html">ModularSet</a> (<a c [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BinaryConv2DAttrs.html">BinaryConv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1SeqStmt_1_1Flattener.html">SeqStmt::Flattener</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSetAnalyzer.html">ModularSetAnalyzer</a> ( [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BinaryDenseAttrs.html">BinaryDenseAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FloatImm.html">FloatImm</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSetNode.html">ModularSetNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BinaryOpNode.html">BinaryOpNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FloatImmNode.html">FloatImmNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Module.html">Module</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)  [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BitPackAttrs.html">BitPackAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorDiv.html">FloorDiv</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ModuleNode.html">ModuleNode</a> (<a class="el" href="namespacetvm_1_1runtime.ht [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Block.html">Block</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorDivNode.html">FloorDivNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Mul.html">Mul</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)    [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1BlockInfo.html">BlockInfo</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorMod.html">FloorMod</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MulNode.html">MulNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)  [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockNode.html">BlockNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorModNode.html">FloorModNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MultiBoxPriorAttrs.html">MultiBoxPriorAttrs</a> (<a class="el" href="namespacetvm_1_1 [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRealize.html">BlockRealize</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStep.html">FollowFusedSplitStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MultiBoxTransformLocAttrs.ht [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRealizeNode.html">BlockRealizeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStepNode.html">FollowFusedSplitStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Muta [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRV.html">BlockRV</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStep.html">FollowSplitStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MutatorNode.html">MutatorNode</a> (<a cl [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRVNode.html">BlockRVNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStepNode.html">FollowSplitStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_n"></a><table border="0" cellspacing="0" ce [...]
-</td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1RunnerInput.html">RunnerInput</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structTVMFuncRegistry.html">TVMFuncRegistry</a>   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockScope.html">BlockScope</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1For.html">For</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1RunnerInputNode.html">RunnerInputNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockScopeNode.html">BlockScopeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ForNode.html">ForNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray.html">NDArray</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::ru [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1Bool.html">Bool</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1FrameBuffer.html">FrameBuffer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1micro__rpc.html">tvm::runtime::micro_rpc</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1NDArrayContainerTrait.html">NDArrayContainerTrait</a> (<a class="el" href= [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Broadcast.html">Broadcast</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1Framer.html">Framer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1micro__rpc.html">tvm::runtime::micro_rpc</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NdarraySizeAttrs.html">NdarraySizeAttrs</a> (< [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1qnn_1_1BroadcastAttrs.html">BroadcastAttrs</a> (<a class="el" href="namespacetvm_1_1relay_1_1qnn.html">tvm::relay::qnn</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ShapeTupleObj_1_1FromStd.html">ShapeTupleObj::FromStd</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NE.html">NE</a> (< [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BroadcastNode.html">BroadcastNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1StringObj_1_1FromStd.html">StringObj::FromStd</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NENode.html">NENode</a> (<a class="el" href="namespace [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1vm_1_1Buffer.html">Buffer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Function.html">Function</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NLLLossAttrs.html">NLLLossAttrs</a> (<a class="el" href="namespa [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Buffer.html">Buffer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionNode.html">FunctionNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1NodeFunctor.html">NodeFunctor</a> (<a class="el" href="namespacetvm.html">tvm</a>)  & [...]
-</td><td valign="top"><a class="el" href="structTVMPackedFunc.html">TVMPackedFunc</a>   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1BufferInfo.html">BufferInfo</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionPattern.html">FunctionPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1NodeFunctor_3_01R_07const_01ObjectRef_01_6n_00_01Args_8_8_8 [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1BufferInfoAnalysis.html">BufferInfoAnalysis</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionPatternNode.html">FunctionPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NonMaximumSuppressionAttr [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1BufferInfoAnalysisNode.html">BufferInfoAnalysisNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FuncType.html">FuncType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NormalAttrs.html">NormalAttrs</a> (<a class="el" href="namespac [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1BufferInfoNode.html">BufferInfoNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FuncTypeNode.html">FuncTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Not.html">Not</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir< [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferLoad.html">BufferLoad</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Fuse.html">Fuse</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NotNode.html">NotNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   < [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferLoadNode.html">BufferLoadNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1FuseNode.html">FuseNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1NullOptType.html">NullOptType</a> (<a class="el" href="namespacetvm_1_1runtime.html" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferNode.html">BufferNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStep.html">FuseStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_o"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><d [...]
-</td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ScatterNDAttrs.html">ScatterNDAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeCall.html">TypeCall</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRealize.html">BufferRealize</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStepNode.html">FuseStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Schedule.html">Schedule</a> (<a class="el" hr [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRealizeNode.html">BufferRealizeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_g"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  g  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjAllocatorBase.html">ObjAllocatorBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Schedule.html">Schedule</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeConstraint.html">TypeConstraint</a> (<a class="el" href="namespacetvm.html" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRegion.html">BufferRegion</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Object.html">Object</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html">ScheduleNode</a> (<a class="el" href="namespacetvm_1_1tir.html" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRegionNode.html">BufferRegionNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GatherAttrs.html">GatherAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectEqual.html">ObjectEqual</a> (<a class="el" href="namespace [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferStore.html">BufferStore</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GatherNDAttrs.html">GatherNDAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectHash.html">ObjectHash</a> (<a class="el" href="namespacetvm_1_1r [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferStoreNode.html">BufferStoreNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GE.html">GE</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjectPtr.html">ObjectPtr</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtim [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Builder.html">Builder</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GenericFunc.html">GenericFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectPtrEqual.html">ObjectPtrEqual</a> (<a class="el" href="namespacetvm_1_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInput.html">BuilderInput</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GenericFuncNode.html">GenericFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectPtrHash.html">ObjectPtrHash</a> (<a class="el" href= [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInputNode.html">BuilderInputNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GENode.html">GENode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">ObjectRef</a> (<a class="el" href= [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderNode.html">BuilderNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GetValidCountsAttrs.html">GetValidCountsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker.html">Ob [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResult.html">BuilderResult</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GlobalPool2DAttrs.html">GlobalPool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker_3_01Arra [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResultNode.html">BuilderResultNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalTypeVar.html">GlobalTypeVar</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker_3_01Map_3_01K_00_01V_01_4_01_4.htm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResult.html">BuildResult</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalTypeVarNode.html">GlobalTypeVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1OnDeviceAttrs.html">OnDeviceAttrs</a> (<a class="el" hr [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResultNode.html">BuildResultNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalVar.html">GlobalVar</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1OneHotAttrs.html">OneHotAttrs</a> (<a class="el" href="namespac [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprPattern.html">ExprPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structMemoryManagerInterface.html">MemoryManagerInterface</a>   </td><td valign="top"><a class="el" href="classtvm_1_1RelayRefTypeNode.html">RelayRefTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a cla [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprPatternNode.html">ExprPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MeshgridAttrs.html">MeshgridAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ReorderStep.html">ReorderStep</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseAttrsNode.html">BaseAttrsNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprRewriter.html">ExprRewriter</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> (<a class="el" href="namespacetvm_1_1runtime_1_1m [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1BaseComputeOpNode.html">BaseComputeOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html">ExprVisitor</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html">MetadataArray</a> (<a class="el" hr [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseExpr.html">BaseExpr</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html">ExprVisitor</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html">MetadataArrayNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1m [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseExprNode.html">BaseExprNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1ExternOp.html">ExternOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a> (<a class="el" href="namespacetvm_1_1runtime_1_1metadata.htm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseFunc.html">BaseFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1ExternOpNode.html">ExternOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html">MetadataBaseNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1meta [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseFuncNode.html">BaseFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTask.html">ExtractedTask</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataNode.html">MetadataNode</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseTensorType.html">BaseTensorType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTaskNode.html">ExtractedTaskNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollector.html">MetricCollector< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseTensorTypeNode.html">BaseTensorTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncObj_1_1Extractor.html">PackedFuncObj::Extractor</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollectorNode.html">MetricC [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseValueEqual.html">BaseValueEqual</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_f"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  f  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Min.html">Min</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ReshapeTensorAttrs.html">ReshapeTensorAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Tuple.html">Tuple</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::re [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1BaseValueHash.html">BaseValueHash</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MinNode.html">MinNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Resize1DAttrs.html">Resize1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>) [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchMatmulAttrs.html">BatchMatmulAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractor.html">FeatureExtractor</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MirrorPadAttrs.html">Mir [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchNormAttrs.html">BatchNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractorNode.html">FeatureExtractorNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BatchToSpaceNDAttrs.html">BatchToSpaceNDAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FeatureSet.html">FeatureSet</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1MixedModeVisitor.html">MixedModeVisitor</a> (<a class="el [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BiasAddAttrs.html">BiasAddAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1FIFOBufferAttrs.html">FIFOBufferAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Mod.html">Mod</a> (<a class="el" href="namespacetvm_1_1tir.htm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BijectiveLayout.html">BijectiveLayout</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1FixedPointMultiplyAttrs.html">FixedPointMultiplyAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ModNode.html">ModNode</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BijectiveLayoutNode.html">BijectiveLayoutNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1SeqStmt_1_1Flattener.html">SeqStmt::Flattener</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSet.html">ModularSet</a> (<a class="el" href= [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BinaryConv2DAttrs.html">BinaryConv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FloatImm.html">FloatImm</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSetAnalyzer.html">ModularSetAnalyzer</a> (<a class="el" href="namespacetvm_1_1arith [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BinaryDenseAttrs.html">BinaryDenseAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FloatImmNode.html">FloatImmNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ModularSetNode.html">ModularSetNode</a> (<a class="el" href="namespacetvm_1_1arith.h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BinaryOpNode.html">BinaryOpNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorDiv.html">FloorDiv</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Module.html">Module</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtim [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1BitPackAttrs.html">BitPackAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorDivNode.html">FloorDivNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ModuleNode.html">ModuleNode</a> (<a class="el" href="namespacetvm_1_1ru [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Block.html">Block</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorMod.html">FloorMod</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Mul.html">Mul</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1BlockInfo.html">BlockInfo</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1FloorModNode.html">FloorModNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1MulNode.html">MulNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)& [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockNode.html">BlockNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStep.html">FollowFusedSplitStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MultiBoxPriorAttrs.html">MultiBoxP [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRealize.html">BlockRealize</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStepNode.html">FollowFusedSplitStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MultiBoxTransformLoc [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRealizeNode.html">BlockRealizeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStep.html">FollowSplitStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Mutator.html">Mutator< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRV.html">BlockRV</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStepNode.html">FollowSplitStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1MutatorNode.html">MutatorNode</a [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockRVNode.html">BlockRVNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1For.html">For</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_n"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  n  </di [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1RunnerInputNode.html">RunnerInputNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structTVMByteArray.html">TVMByteArray</a>   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockScope.html">BlockScope</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ForNode.html">ForNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1RunnerNode.html">RunnerNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BlockScopeNode.html">BlockScopeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1FrameBuffer.html">FrameBuffer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1micro__rpc.html">tvm::runtime::micro_rpc</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray.html">NDArray</a> [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1Bool.html">Bool</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1Framer.html">Framer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1micro__rpc.html">tvm::runtime::micro_rpc</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1NDArrayContainerTrait.html">NDArrayContainerTrait</a> (<a class="el" href="namespace [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Broadcast.html">Broadcast</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1StringObj_1_1FromStd.html">StringObj::FromStd</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NdarraySizeAttrs.html">NdarraySizeAttrs</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1qnn_1_1BroadcastAttrs.html">BroadcastAttrs</a> (<a class="el" href="namespacetvm_1_1relay_1_1qnn.html">tvm::relay::qnn</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ShapeTupleObj_1_1FromStd.html">ShapeTupleObj::FromStd</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NE.html">NE</a> (< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BroadcastNode.html">BroadcastNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Function.html">Function</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NENode.html">NENode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1vm_1_1Buffer.html">Buffer</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionNode.html">FunctionNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NLLLossAttrs.html">NLLLossAttrs</a> (<a class="el" href= [...]
+</td><td valign="top"><a class="el" href="structTVMMutableFuncRegistry.html">TVMMutableFuncRegistry</a>   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Buffer.html">Buffer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionPattern.html">FunctionPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1NodeFunctor.html">NodeFunctor</a> (<a class="el" href="namespacetvm.html">tvm</a>) & [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1BufferInfo.html">BufferInfo</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1FunctionPatternNode.html">FunctionPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1NodeFunctor_3_01R_07const_01ObjectRef_01_6n_00_01Ar [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1BufferInfoAnalysis.html">BufferInfoAnalysis</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FuncType.html">FuncType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NonMaximumSuppressionAttrs.html">NonMaximumSuppressionAttrs</a> (<a clas [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1BufferInfoAnalysisNode.html">BufferInfoAnalysisNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1FuncTypeNode.html">FuncTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1NormalAttrs.html">NormalAttrs</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1BufferInfoNode.html">BufferInfoNode</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp.html">tvm::tir::usmp</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Fuse.html">Fuse</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Not.html">Not</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferLoad.html">BufferLoad</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1FuseNode.html">FuseNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1NotNode.html">NotNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferLoadNode.html">BufferLoadNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStep.html">FuseStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1NullOptType.html">NullOptType</a> (<a class="e [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferNode.html">BufferNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStepNode.html">FuseStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_o"></a><table border="0" cellspacing="0" cellpadding="0"><t [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Schedule.html">Schedule</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Type.html">Type</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRealize.html">BufferRealize</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_g"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  g  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Schedule.html">Schedule</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeCall.html">TypeCall</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRealizeNode.html">BufferRealizeNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjAllocatorBase.html">ObjAllocatorBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html">ScheduleNode</a> (<a class="el" h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRegion.html">BufferRegion</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GatherAttrs.html">GatherAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Object.html">Object</a> (<a class="el" href="namespacetvm_1_1runtime.html [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferRegionNode.html">BufferRegionNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GatherNDAttrs.html">GatherNDAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectEqual.html">ObjectEqual</a> (<a class="el" href="names [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferStore.html">BufferStore</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GE.html">GE</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectHash.html">ObjectHash</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a> [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1BufferStoreNode.html">BufferStoreNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GenericFunc.html">GenericFunc</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjectPtr.html">ObjectPtr</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Builder.html">Builder</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GenericFuncNode.html">GenericFuncNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectPtrEqual.html">ObjectPtrEqual</a> (<a class="el" href="namespa [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInput.html">BuilderInput</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GENode.html">GENode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectPtrHash.html">ObjectPtrHash</a> (<a class="el" href [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInputNode.html">BuilderInputNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GetValidCountsAttrs.html">GetValidCountsAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">O [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderNode.html">BuilderNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GlobalPool2DAttrs.html">GlobalPool2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker.html">Object [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResult.html">BuilderResult</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalTypeVar.html">GlobalTypeVar</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker_3_01Array_3_01T_01_4_01_4.html">ObjectType [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResultNode.html">BuilderResultNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalTypeVarNode.html">GlobalTypeVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1ObjectTypeChecker_3_01Map_3_01K_00_01V_01_4_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResult.html">BuildResult</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalVar.html">GlobalVar</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1OnDeviceAttrs.html">OnDeviceAttrs</a> (<a class="el" href="namespacetvm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResultNode.html">BuildResultNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1GlobalVarNode.html">GlobalVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1OneHotAttrs.html">OneHotAttrs</a> (<a class="el" href=" [...]
<tr><td rowspan="2" valign="bottom"><a name="letter_c"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  c  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1GlobalVarNode.html">GlobalVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Op.html">Op</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1SearchStrategy.html">SearchStrategy</a> (<a class="el" href="namespacetvm_1_1meta__schedule.html">tvm::meta_schedule</a>)  [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1algo_1_1GreedyBase.html">GreedyBase</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp_1_1algo.html">tvm::tir::usmp::algo</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1OpAttrMap.html">OpAttrMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1SearchStrategyNode.html">SearchStrategyNode</a> (<a clas [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStep.html">CacheReadStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GridSampleAttrs.html">GridSampleAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Operation.html">Operation</a> (<a [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStepNode.html">CacheReadStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GroupNormAttrs.html">GroupNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1OperationNode.html">Operati [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStep.html">CacheWriteStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GT.html">GT</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpImplementation.html">OpImplementation</a> (<a class="el" hr [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStepNode.html">CacheWriteStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GTNode.html">GTNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpImplementationNode.html">OpImplementationNo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Call.html">Call</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_h"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  h  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1OpNode.html">OpNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1SelectSEqualReduce.html">SelectSEqualReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1TypeName_3_01void_01_5_01_4.html">TypeName< void * ></a> (<a class="el" href="na [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Call.html">Call</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1OpRegEntry.html">OpRegEntry</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1SelectSEqualReduce_3_01T_00_01TraitName_00_01false_01_4.html">SelectSEqualReduce< T, TraitName, false ></a> (<a [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1profiling_1_1CallFrame.html">CallFrame</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1SEqualReducer_1_1Handler.html">SEqualReducer::Handler</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpSpecialization.html">OpSpeciali [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CallLoweredAttrs.html">CallLoweredAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1SHashReducer_1_1Handler.html">SHashReducer::Handler</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpSpecializationNode.html">OpSpecializationNode</a> (<a class="e [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallNode.html">CallNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1SimpleObjAllocator_1_1Handler.html">SimpleObjAllocator::Handler</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpStrategy.html">OpStrategy</a> (<a cl [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CallNode.html">CallNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structdmlc_1_1serializer_1_1Handler_3_01DLDataType_01_4.html">Handler< DLDataType ></a> (<a class="el" href="namespacedmlc_1_1serializer.html">dmlc::serializer</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpStrategyNode.html">OpStrategyNo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallPattern.html">CallPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structdmlc_1_1serializer_1_1Handler_3_01DLDevice_01_4.html">Handler< DLDevice ></a> (<a class="el" href="namespacedmlc_1_1serializer.html">dmlc::serializer</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Optional.html">Optional [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallPatternNode.html">CallPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParams.html">HardwareParams</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Or.html">Or</a> (<a class="el" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1CanonicalSimplifier.html">CanonicalSimplifier</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParamsNode.html">HardwareParamsNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1OrNode.html">Or [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Cast.html">Cast</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1HybridOp.html">HybridOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_p"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  p  </div></td> [...]
-</td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1SequenceMaskAttrs.html">SequenceMaskAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeVarNode.html">TypeVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CastAttrs.html">CastAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1HybridOpNode.html">HybridOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1Sequential.html">Sequential</a> (<a class="el" href="namespacetvm_1_1transform [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CastHintAttrs.html">CastHintAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_i"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  i  </div></td></tr></table>
-</td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFunc.html">PackedFunc</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1SequentialNode.html">SequentialNode</a> (<a class="el" href="namespacetvm_1_1transform.html">tvm::transform</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_u"></a><table border="0" cellspacing="0" cellpadding="0"><t [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1algo_1_1GreedyBase.html">GreedyBase</a> (<a class="el" href="namespacetvm_1_1tir_1_1usmp_1_1algo.html">tvm::tir::usmp::algo</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Op.html">Op</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1SearchStrategyNode.html">SearchStrategyNode</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GridSampleAttrs.html">GridSampleAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1OpAttrMap.html">OpAttrMap</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1SearchTask.html">SearchTask</a> (<a class="el" href="namespacetvm_1_1auto__schedul [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStep.html">CacheReadStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1GroupNormAttrs.html">GroupNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1Operation.html">Operation</a> (<a c [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStepNode.html">CacheReadStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GT.html">GT</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1OperationNode.html">OperationNode</a> (<a class="el" href= [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStep.html">CacheWriteStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1GTNode.html">GTNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpImplementation.html">OpImplementation</a> (<a class [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStepNode.html">CacheWriteStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_h"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  h  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpImplementationNode.html">OpImplementationNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1SelectSEqualReduce.html">SelectSEqualReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1TypeName_3_01int64__t_01_4.html"> [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Call.html">Call</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1OpNode.html">OpNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1SelectSEqualReduce_3_01T_00_01TraitName_00_01false_01_4.html">SelectSEqualReduce< T, TraitName, false ></a> (<a c [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Call.html">Call</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1SimpleObjAllocator_1_1Handler.html">SimpleObjAllocator::Handler</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1OpRegEntry.html">OpRegEntry</a> (<a class="el" href="namespac [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1profiling_1_1CallFrame.html">CallFrame</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::profiling</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1SEqualReducer_1_1Handler.html">SEqualReducer::Handler</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpSpecialization.html">OpSpeciali [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CallLoweredAttrs.html">CallLoweredAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1SHashReducer_1_1Handler.html">SHashReducer::Handler</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpSpecializationNode.html">OpSpecializationNode</a> (<a class="e [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallNode.html">CallNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structdmlc_1_1serializer_1_1Handler_3_01DLDataType_01_4.html">Handler< DLDataType ></a> (<a class="el" href="namespacedmlc_1_1serializer.html">dmlc::serializer</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpStrategy.html">OpStrategy [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CallNode.html">CallNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structdmlc_1_1serializer_1_1Handler_3_01DLDevice_01_4.html">Handler< DLDevice ></a> (<a class="el" href="namespacedmlc_1_1serializer.html">dmlc::serializer</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1OpStrategyNode.html">OpStrategyNode</ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallPattern.html">CallPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParams.html">HardwareParams</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Optional.html">Optional</a> (<a cla [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1CallPatternNode.html">CallPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParamsNode.html">HardwareParamsNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Or.html">Or</a> (<a cla [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1CanonicalSimplifier.html">CanonicalSimplifier</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1HybridOp.html">HybridOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1OrNode.html">OrNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tv [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Cast.html">Cast</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1HybridOpNode.html">HybridOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_p"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  p  </d [...]
+</td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1Sequential.html">Sequential</a> (<a class="el" href="namespacetvm_1_1transform.html">tvm::transform</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeVar.html">TypeVar</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CastAttrs.html">CastAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td rowspan="2" valign="bottom"><a name="letter_i"></a><table border="0" cellspacing="0" cellpadding="0"><tr><td><div class="ah">  i  </div></td></tr></table>
+</td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1SequentialNode.html">SequentialNode</a> (<a class="el" href="namespacetvm_1_1transform.html">tvm::transform</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1TypeVarNode.html">TypeVarNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td></tr>
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CastHintAttrs.html">CastHintAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFunc.html">PackedFunc</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1Session.html">Session</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CastNode.html">CastNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Id.html">Id</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFuncObj.html">PackedFuncObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtim [...]
</td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CastNode.html">CastNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFuncObj.html">PackedFuncObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1Session.html">Session</a> (<a class="el" href="namespacetv [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Clause.html">Clause</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Id.html">Id</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFuncSubObj.html">PackedFuncSubObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ClauseNode.html">ClauseNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IdNode.html">IdNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter.html">PackedFuncValueConverter</a> (<a class="el" href=" [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ClipAttrs.html">ClipAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1If.html">If</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01Optional_3_01T_01_4_01_4.html">PackedFuncValueConverter< O [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Closure.html">Closure</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfNode.html">IfNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01PrimExpr_01_4.html">PackedFuncValueConverter< Pri [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ClosureObj.html">ClosureObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfPattern.html">IfPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01tvm_1_1Bool_01_4.html">PackedFuncValueCo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CmpOpNode.html">CmpOpNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfPatternNode.html">IfPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01tvm_1_1Integer_01_4.html">PackedFuncValueConve [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Clause.html">Clause</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IdNode.html">IdNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1PackedFuncSubObj.html">PackedFuncSubObj</a> (<a class="el" href="namespacetvm_1_1runtime.h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ClauseNode.html">ClauseNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1If.html">If</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter.html">PackedFuncValueConverter</a> (<a class="el" href="namespac [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ClipAttrs.html">ClipAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfNode.html">IfNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01Optional_3_01T_01_4_01_4.html">PackedFuncValueConvert [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1Closure.html">Closure</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfPattern.html">IfPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01PrimExpr_01_4.html">PackedFuncValueConverter&l [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1ClosureObj.html">ClosureObj</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1IfPatternNode.html">IfPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01tvm_1_1Bool_01_4.html">PackedFun [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CmpOpNode.html">CmpOpNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IfThenElse.html">IfThenElse</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_01tvm_1_1Integer_01_4.html">PackedFuncValueConverter< tvm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CommReducer.html">CommReducer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IfThenElseNode.html">IfThenElseNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_1_1tvm_1_1runtime_1_1String_01_4.html">PackedFun [...]
</td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CommReducer.html">CommReducer</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IfThenElse.html">IfThenElse</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1PackedFuncValueConverter_3_1_1tvm_1_1runtime_1_1String_01_4.html">PackedFuncValueCo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CommReducerNode.html">CommReducerNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IfThenElseNode.html">IfThenElseNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1PacketFieldSizeBytes.html">PacketFieldSizeBytes</a> (< [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1CompilationConfig.html">CompilationConfig</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSEqualReduce.html">ImplSEqualReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1PadAttrs.html">PadAttrs</a> (<a class="el" href="namespacetvm_1_1relay [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1CompilationConfigNode.html">CompilationConfigNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSEqualReduce_3_01T_00_01true_01_4.html">ImplSEqualReduce< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1Pass.html">Pass</a> ( [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1CompileError.html">CompileError</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSHashReduce.html">ImplSHashReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassContext.html">PassContext</a> (<a class="el" href="namespacetvm_1_1transfor [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CompilerAttrs.html">CompilerAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSHashReduce_3_01T_00_01true_01_4.html">ImplSHashReduce< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassContextNo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStep.html">ComputeAtStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplVisitAttrs.html">ImplVisitAttrs</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassInfo.html">PassInfo</ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStepNode.html">ComputeAtStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplVisitAttrs_3_01T_00_01true_01_4.html">ImplVisitAttrs< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="clas [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAG.html">ComputeDAG</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IncompleteType.html">IncompleteType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1instrument_1_1PassInstrument.html">PassInstrument</a> (<a class="el" href [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAGNode.html">ComputeDAGNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IncompleteTypeNode.html">IncompleteTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1instrument_1_1PassInstrumentNode.html">PassInstrumentNode [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStep.html">ComputeInlineStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IndexMap.html">IndexMap</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassNode.html">PassNode</a> (<a class=" [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStepNode.html">ComputeInlineStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IndexMapNode.html">IndexMapNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Pattern.html">Pattern</a> ( [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1ComputeOp.html">ComputeOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1InitOpAttrs.html">InitOpAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternConstructor.html">PatternConstructor</a> (<a class="el" href="namespacetvm_1_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1ComputeOpNode.html">ComputeOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1InplaceArrayBase.html">InplaceArrayBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternConstructorNode.html">PatternConstructorNode</a> (<a c [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStep.html">ComputeRootStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1InstanceNormAttrs.html">InstanceNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternFunctor.html">P [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStepNode.html">ComputeRootStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1vm_1_1Instruction.html">Instruction</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Patt [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConcatenateAttrs.html">ConcatenateAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Instruction.html">Instruction</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternMutator.html">PatternMutator</a> (<a class="el" href="namesp [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Constant.html">Constant</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKind.html">InstructionKind</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternNode.html">PatternNode</a> (<a class="el" href="namespacetvm_1_1relay [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantNode.html">ConstantNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKindNode.html">InstructionKindNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternTuple.html">PatternTuple</a> (<a class="el" href="nam [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantPattern.html">ConstantPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKindRegEntry.html">InstructionKindRegEntry</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternTupleNode.html">PatternTupleNode</a> (< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1tir_1_1CommReducerNode.html">CommReducerNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSEqualReduce.html">ImplSEqualReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1micro__rpc_1_1PacketFieldSizeBytes.html">PacketFieldSi [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1CompilationConfig.html">CompilationConfig</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSEqualReduce_3_01T_00_01true_01_4.html">ImplSEqualReduce< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1PadAttrs.html">PadAttrs</a> (<a [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1CompilationConfigNode.html">CompilationConfigNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSHashReduce.html">ImplSHashReduce</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1Pass.html">Pass</a> (<a class="el" href="namespacetvm_1_1tran [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1CompileError.html">CompileError</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplSHashReduce_3_01T_00_01true_01_4.html">ImplSHashReduce< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassContext.html">PassContext</a> (<a cla [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1CompilerAttrs.html">CompilerAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplVisitAttrs.html">ImplVisitAttrs</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassContextNode.html">PassContextNode</a> (<a class=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStep.html">ComputeAtStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1ImplVisitAttrs_3_01T_00_01true_01_4.html">ImplVisitAttrs< T, true ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1 [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStepNode.html">ComputeAtStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IncompleteType.html">IncompleteType</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassInfoNode.html">PassInfoNode</a> (<a class= [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAG.html">ComputeDAG</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IncompleteTypeNode.html">IncompleteTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1instrument_1_1PassInstrument.html">PassInstrument</a> (<a class=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAGNode.html">ComputeDAGNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IndexMap.html">IndexMap</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1instrument_1_1PassInstrumentNode.html">PassInstrumentNode< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStep.html">ComputeInlineStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1IndexMapNode.html">IndexMapNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1transform_1_1PassNode.html">PassNode</a> (<a [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStepNode.html">ComputeInlineStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1InitOpAttrs.html">InitOpAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Pattern.html">Pattern< [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1ComputeOp.html">ComputeOp</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1InplaceArrayBase.html">InplaceArrayBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternConstructor.html">PatternConstructor</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1te_1_1ComputeOpNode.html">ComputeOpNode</a> (<a class="el" href="namespacetvm_1_1te.html">tvm::te</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1InstanceNormAttrs.html">InstanceNormAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternConstructorNode.html">PatternConstructorNode</a> (<a clas [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStep.html">ComputeRootStep</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1runtime_1_1vm_1_1Instruction.html">Instruction</a> (<a class="el" href="namespacetvm_1_1runtime_1_1vm.html">tvm::runtime::vm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternFunct [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStepNode.html">ComputeRootStepNode</a> (<a class="el" href="namespacetvm_1_1auto__scheduler.html">tvm::auto_scheduler</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Instruction.html">Instruction</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pa [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConcatenateAttrs.html">ConcatenateAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKind.html">InstructionKind</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternMutator.html">PatternMutator</a> (<a class="el" href [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1Constant.html">Constant</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKindNode.html">InstructionKindNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternNode.html">PatternNode</a> (<a class="el" href="namespacetvm_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantNode.html">ConstantNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionKindRegEntry.html">InstructionKindRegEntry</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternTuple.html">PatternTuple</a> (<a class="el" h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantPattern.html">ConstantPattern</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionNode.html">InstructionNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternTupleNode.html">PatternTupleNode</a> (<a class="el" hre [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantPatternNode.html">ConstantPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraints.html">IntConstraints</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVar.html">PatternVar</a> (<a class="el" hre [...]
</td></tr>
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstantPatternNode.html">ConstantPatternNode</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1InstructionNode.html">InstructionNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVar.html">PatternVar</a> (<a class="el" href="n [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBound.html">ConstIntBound</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraints.html">IntConstraints</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVarNode.html">PatternVarNode</a> (<a class="el" href="n [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBoundAnalyzer.html">ConstIntBoundAnalyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsNode.html">IntConstraintsNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVisitor.html">PatternVisitor</a [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBoundNode.html">ConstIntBoundNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransform.html">IntConstraintsTransform</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternWildcard.html">PatternWildcar [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstraintContext.html">ConstraintContext</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransformNode.html">IntConstraintsTransformNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternWildcardNode.html">Pa [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1Constructor.html">Constructor</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Integer.html">Integer</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1PercentNode.html">PercentNode</a> (<a class="el" href="namespacetvm_1_1runtime_1_1profiling.html">tvm::runtime::pro [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1ConstructorNode.html">ConstructorNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1InterpreterClosure.html">InterpreterClosure</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1PlaceholderOp.html">PlaceholderOp</a> (<a class="el" href="namespacetvm_1_1te. [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstructorValue.html">ConstructorValue</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1InterpreterClosureObj.html">InterpreterClosureObj</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1PlaceholderOpNode.html">PlaceholderOpNode</a> [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConstructorValueObj.html">ConstructorValueObj</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntGroupBounds.html">IntGroupBounds</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PointerType.html">PointerType</a> (<a class="el" href="nam [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray_1_1Container.html">NDArray::Container</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntGroupBoundsNode.html">IntGroupBoundsNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PointerTypeNode.html">PointerTypeNode</a> (<a [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray_1_1ContainerBase.html">NDArray::ContainerBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IntImm.html">IntImm</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1PoolAllocation.html">PoolAllocation</a> (<a class="el" href="namespacet [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv1DAttrs.html">Conv1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IntImmNode.html">IntImmNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1PoolAllocationNode.html">PoolAllocationNode</a> (<a class="el" href="namespacetvm_1_1tir_1_ [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv1DTransposeAttrs.html">Conv1DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSet.html">IntSet</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PoolInfo.html">PoolInfo</a> (<a class="el" href="namespacetvm.html">tvm< [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DAttrs.html">Conv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSetAnalyzer.html">IntSetAnalyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1PoolInfoNode.html">PoolInfoNode</a> (<a class="el" href="namespacetvm.htm [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBound.html">ConstIntBound</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsNode.html">IntConstraintsNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVarNode.html">PatternVarNode</a> (<a class="el" [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBoundAnalyzer.html">ConstIntBoundAnalyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransform.html">IntConstraintsTransform</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternVisitor.html">Pattern [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstIntBoundNode.html">ConstIntBoundNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransformNode.html">IntConstraintsTransformNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternWildcard.html">Patter [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1arith_1_1ConstraintContext.html">ConstraintContext</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1Integer.html">Integer</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternWildcardNode.html">PatternWildcardNode</a> (<a class="el" href="namespacetvm_1_1relay. [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1Constructor.html">Constructor</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1InterpreterClosure.html">InterpreterClosure</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1PercentNode.html">PercentNode</a> (<a class="el" href="namespacetvm_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1ConstructorNode.html">ConstructorNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1InterpreterClosureObj.html">InterpreterClosureObj</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1PlaceholderOp.html">PlaceholderOp</a> (<a class="el" href="namespacetvm_ [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1relay_1_1ConstructorValue.html">ConstructorValue</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntGroupBounds.html">IntGroupBounds</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1PlaceholderOpNode.html">PlaceholderOpNode</a> (<a class="el [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1ConstructorValueObj.html">ConstructorValueObj</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntGroupBoundsNode.html">IntGroupBoundsNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PointerType.html">PointerType</a> (<a class="el" h [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray_1_1Container.html">NDArray::Container</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IntImm.html">IntImm</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PointerTypeNode.html">PointerTypeNode</a> (<a class="el" href="namespacetvm.html">tvm</a>) [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1NDArray_1_1ContainerBase.html">NDArray::ContainerBase</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IntImmNode.html">IntImmNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1usmp_1_1PoolAllocation.html">PoolAllocation</a> (<a class="el" href="na [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv1DAttrs.html">Conv1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSet.html">IntSet</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1tir_1_1usmp_1_1PoolAllocationNode.html">PoolAllocationNode</a> (<a class="el" href="names [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv1DTransposeAttrs.html">Conv1DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSetAnalyzer.html">IntSetAnalyzer</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1PoolInfo.html">PoolInfo</a> (<a class="el" href="namespa [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DAttrs.html">Conv2DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSetNode.html">IntSetNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1PoolInfoNode.html">PoolInfoNode</a> (<a class="el" href="namespacetvm.html">tvm</ [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DTransposeAttrs.html">Conv2DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IRModule.html">IRModule</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Postproc.html">Postproc</a> (<a class="el" href="namespacetvm_1_1meta__sche [...]
</td></tr>
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DTransposeAttrs.html">Conv2DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1arith_1_1IntSetNode.html">IntSetNode</a> (<a class="el" href="namespacetvm_1_1arith.html">tvm::arith</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1Postproc.html">Postproc</a> (<a class="el" hre [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DWinogradAttrs.html">Conv2DWinogradAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IRModule.html">IRModule</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1PostprocNode.html">PostprocNode</a> (<a class="el" href="namespacetvm_1_1meta [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DWinogradNNPACKWeightTransformAttrs.html">Conv2DWinogradNNPACKWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IRModuleNode.html">IRModuleNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStep.html">PragmaSte [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DAttrs.html">Conv3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1is__specialized.html">is_specialized</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStepNode.html">PragmaStepNode</a> (<a class= [...]
-<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DTransposeAttrs.html">Conv3DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1is__specialized_3_01Container_3_01Args_8_8_8_01_4_00_01Container_01_4.html">is_specialized< Container< Args... >, Container ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DWinogradAttrs.html">Conv2DWinogradAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1IRModuleNode.html">IRModuleNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1meta__schedule_1_1PostprocNode.html">PostprocNode</a> (<a class="el" href="namespacetvm [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv2DWinogradNNPACKWeightTransformAttrs.html">Conv2DWinogradNNPACKWeightTransformAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1is__specialized.html">is_specialized</a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1auto__sched [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DAttrs.html">Conv3DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1detail_1_1is__specialized_3_01Container_3_01Args_8_8_8_01_4_00_01Container_01_4.html">is_specialized< Container< Args... >, Container ></a> (<a class="el" href="namespacetvm_1_1detail.html">tvm::detail</a>)   </td><td valign="top"> [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DTransposeAttrs.html">Conv3DTransposeAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1IterAdapter.html">IterAdapter</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Prefetch.html">Prefetch</a> (<a class="el" href=" [...]
+<tr><td valign="top"><a class="el" href="structtvm_1_1relay_1_1Conv3DWinogradAttrs.html">Conv3DWinogradAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1runtime_1_1MapNode_1_1iterator.html">MapNode::iterator</a> (<a class="el" href="namespacetvm_1_1runtime.html">tvm::runtime</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1PrefetchNode.html">PrefetchNode</a> ( [...]
<tr><td></td><td></td><td></td><td></td><td></td></tr>
</table>
<div class="qindex"><a class="qindex" href="#letter_a">a</a> | <a class="qindex" href="#letter_b">b</a> | <a class="qindex" href="#letter_c">c</a> | <a class="qindex" href="#letter_d">d</a> | <a class="qindex" href="#letter_e">e</a> | <a class="qindex" href="#letter_f">f</a> | <a class="qindex" href="#letter_g">g</a> | <a class="qindex" href="#letter_h">h</a> | <a class="qindex" href="#letter_i">i</a> |& [...]
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef__coll__graph.svg b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef__coll__graph.svg
index b304ecc87..46e398b26 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef__coll__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef__coll__graph.svg
@@ -9,9 +9,9 @@
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 453)">
<title>tvm::runtime::ObjectRef</title>
<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-453 144,-453 144,4 -4,4"/>
-<!-- Node391 -->
+<!-- Node393 -->
<g id="node1" class="node">
-<title>Node391</title>
+<title>Node393</title>
<polygon fill="#bfbfbf" stroke="#000000" points="3,-.5 3,-222.5 137,-222.5 137,-.5 3,-.5"/>
<text text-anchor="middle" x="70" y="-210.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectRef</text>
<polyline fill="none" stroke="#000000" points="3,-203.5 137,-203.5 "/>
@@ -34,9 +34,9 @@
<text text-anchor="start" x="11" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># FFIClearAfterMove()</text>
<text text-anchor="start" x="11" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetDataPtr()</text>
</g>
-<!-- Node392 -->
+<!-- Node394 -->
<g id="node2" class="node">
-<title>Node392</title>
+<title>Node394</title>
<g id="a_node2"><a xlink:href="classtvm_1_1runtime_1_1ObjectPtr.html" target="_top" xlink:title="{tvm::runtime::ObjectPtr\l\< tvm::runtime::Object \>\n||+ ObjectPtr()\l+ ObjectPtr()\l+ ObjectPtr()\l+ ObjectPtr()\l+ ObjectPtr()\l+ ObjectPtr()\l+ ~ObjectPtr()\l+ swap()\l+ get()\l+ operator-\>()\land 11 more...\l}">
<polygon fill="#ffffff" stroke="#000000" points="0,-270.5 0,-448.5 140,-448.5 140,-270.5 0,-270.5"/>
<text text-anchor="start" x="8" y="-436.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectPtr</text>
@@ -58,9 +58,9 @@
</a>
</g>
</g>
-<!-- Node392->Node391 -->
+<!-- Node394->Node393 -->
<g id="edge1" class="edge">
-<title>Node392->Node391</title>
+<title>Node394->Node393</title>
<path fill="none" stroke="#404040" d="M70,-270.3167C70,-258.8765 70,-247.0062 70,-235.1402"/>
<polygon fill="none" stroke="#404040" points="70.0001,-234.7944 66,-228.7944 70,-222.7944 74,-228.7943 70.0001,-234.7944"/>
<text text-anchor="middle" x="89.5" y="-244" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #data_</text>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1Object__coll__graph.svg b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1Object__coll__graph.svg
index ce6728003..7345dbcb5 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1Object__coll__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1Object__coll__graph.svg
@@ -9,9 +9,9 @@
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 392)">
<title>tvm::runtime::Object</title>
<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-392 257,-392 257,4 -4,4"/>
-<!-- Node607 -->
+<!-- Node609 -->
<g id="node1" class="node">
-<title>Node607</title>
+<title>Node609</title>
<polygon fill="#bfbfbf" stroke="#000000" points="0,-.5 0,-387.5 183,-387.5 183,-.5 0,-.5"/>
<text text-anchor="middle" x="91.5" y="-375.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
<polyline fill="none" stroke="#000000" points="0,-368.5 183,-368.5 "/>
@@ -49,9 +49,9 @@
<text text-anchor="start" x="8" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DecRef()</text>
<text text-anchor="start" x="8" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetOrAllocRuntimeTypeIndex()</text>
</g>
-<!-- Node607->Node607 -->
+<!-- Node609->Node609 -->
<g id="edge1" class="edge">
-<title>Node607->Node607</title>
+<title>Node609->Node609</title>
<path fill="none" stroke="#404040" d="M183.3625,-256.0888C194.0482,-244.6299 201,-223.9336 201,-194 201,-171.3159 197.0077,-153.9367 190.4236,-141.8623"/>
<polygon fill="none" stroke="#404040" points="190.3069,-141.6977 183.5725,-139.1192 183.3625,-131.9112 190.0969,-134.4897 190.3069,-141.6977"/>
<text text-anchor="middle" x="227" y="-191.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #deleter_</text>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html
index 716ef083e..2650b9265 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html
@@ -71,7 +71,7 @@ $(function() {
</div><!--header-->
<div class="contents">
-<p>A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a> fields with complex elements. These array fields should be accessed from C++ using the Metadata wrapper classes. This class lazily instantiates those wrappers as they are accessed.
+<p>A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a> fields with complex elements. These array fields should be accessed from C++ using the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> wrapper classes. This class lazily instantiates those wrappers as they are accessed.
<a href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html#details">More...</a></p>
<p><code>#include <<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a>></code></p>
@@ -109,7 +109,7 @@ Public Member Functions</h2></td></tr>
<div class="textblock"><h3>template<typename C, class Ref><br />
class tvm::runtime::metadata::ArrayAccessor< C, Ref ></h3>
-<p>A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a> fields with complex elements. These array fields should be accessed from C++ using the Metadata wrapper classes. This class lazily instantiates those wrappers as they are accessed. </p>
+<p>A span-like class which permits access to <a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a> fields with complex elements. These array fields should be accessed from C++ using the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> wrapper classes. This class lazily instantiates those wrappers as they are accessed. </p>
</div><h2 class="groupheader">Member Typedef Documentation</h2>
<a id="a3904284b75894a8ec7dc385a708edec3"></a>
<h2 class="memtitle"><span class="permalink"><a href="#a3904284b75894a8ec7dc385a708edec3">◆ </a></span>const_iterator</h2>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html
index 6733c6249..6c4eddfa9 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html
@@ -71,7 +71,7 @@ $(function() {
</div><!--header-->
<div class="contents">
-<p>A specialization of <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" title="A span-like class which permits access to Array fields with complex elements. These array fields shou...">ArrayAccessor</a> for <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a>. This class is needed because the <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a> constructor sig [...]
+<p>A specialization of <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" title="A span-like class which permits access to Array fields with complex elements. These array fields shou...">ArrayAccessor</a> for <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a>. This class is needed because the <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a> constructor sig [...]
<a href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor_3_01const_01char_01_5_00_01_1_1tvm_1_1runtime_1_1String_01_4.html#details">More...</a></p>
<p><code>#include <<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a>></code></p>
@@ -108,7 +108,7 @@ Public Member Functions</h2></td></tr>
<div class="textblock"><h3>template<><br />
class tvm::runtime::metadata::ArrayAccessor< const char *, ::tvm::runtime::String ></h3>
-<p>A specialization of <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" title="A span-like class which permits access to Array fields with complex elements. These array fields shou...">ArrayAccessor</a> for <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a>. This class is needed because the <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a> constructor sig [...]
+<p>A specialization of <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1ArrayAccessor.html" title="A span-like class which permits access to Array fields with complex elements. These array fields shou...">ArrayAccessor</a> for <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a>. This class is needed because the <a class="el" href="classtvm_1_1runtime_1_1String.html" title="Reference to string objects. ">String</a> constructor sig [...]
</div><h2 class="groupheader">Member Typedef Documentation</h2>
<a id="a7068f394277484b87b066cd34979d03f"></a>
<h2 class="memtitle"><span class="permalink"><a href="#a7068f394277484b87b066cd34979d03f">◆ </a></span>const_iterator</h2>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html
index d378b929b..7c13f1a7b 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html
@@ -70,7 +70,7 @@ $(function() {
</div><!--header-->
<div class="contents">
-<p>An iterator implementation that lazily instantiates the C++ wrapping Metadata class.
+<p>An iterator implementation that lazily instantiates the C++ wrapping <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> class.
<a href="classtvm_1_1runtime_1_1metadata_1_1ArrayIterator.html#details">More...</a></p>
<p><code>#include <<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a>></code></p>
@@ -98,7 +98,7 @@ Public Member Functions</h2></td></tr>
<div class="textblock"><h3>template<typename C, class Ref><br />
class tvm::runtime::metadata::ArrayIterator< C, Ref ></h3>
-<p>An iterator implementation that lazily instantiates the C++ wrapping Metadata class. </p>
+<p>An iterator implementation that lazily instantiates the C++ wrapping <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> class. </p>
</div><h2 class="groupheader">Constructor & Destructor Documentation</h2>
<a id="adf1e3e2969deb955e49e6f93e373a11e"></a>
<h2 class="memtitle"><span class="permalink"><a href="#adf1e3e2969deb955e49e6f93e373a11e">◆ </a></span>ArrayIterator()</h2>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata-members.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata-members.html
new file mode 100644
index 000000000..6864b7203
--- /dev/null
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata-members.html
@@ -0,0 +1,102 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml">
+<head>
+<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
+<meta http-equiv="X-UA-Compatible" content="IE=9"/>
+<meta name="generator" content="Doxygen 1.8.13"/>
+<meta name="viewport" content="width=device-width, initial-scale=1"/>
+<title>tvm: Member List</title>
+<link href="tabs.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="jquery.js"></script>
+<script type="text/javascript" src="dynsections.js"></script>
+<link href="search/search.css" rel="stylesheet" type="text/css"/>
+<script type="text/javascript" src="search/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+</head>
+<body>
+<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
+<div id="titlearea">
+<table cellspacing="0" cellpadding="0">
+ <tbody>
+ <tr style="height: 56px;">
+ <td id="projectalign" style="padding-left: 0.5em;">
+ <div id="projectname">tvm
+ </div>
+ </td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.13 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+<script type="text/javascript" src="menudata.js"></script>
+<script type="text/javascript" src="menu.js"></script>
+<script type="text/javascript">
+$(function() {
+ initMenu('',true,false,'search.php','Search');
+ $(document).ready(function() { init_search(); });
+});
+</script>
+<div id="main-nav"></div>
+<!-- window showing the filter options -->
+<div id="MSearchSelectWindow"
+ onmouseover="return searchBox.OnSearchSelectShow()"
+ onmouseout="return searchBox.OnSearchSelectHide()"
+ onkeydown="return searchBox.OnSearchSelectKey(event)">
+</div>
+
+<!-- iframe showing the search results (closed by default) -->
+<div id="MSearchResultsWindow">
+<iframe src="javascript:void(0)" frameborder="0"
+ name="MSearchResults" id="MSearchResults">
+</iframe>
+</div>
+
+<div id="nav-path" class="navpath">
+ <ul>
+<li class="navelem"><a class="el" href="namespacetvm.html">tvm</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime.html">runtime</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">metadata</a></li><li class="navelem"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a></li> </ul>
+</div>
+</div><!-- top -->
+<div class="header">
+ <div class="headertitle">
+<div class="title">tvm::runtime::metadata::Metadata Member List</div> </div>
+</div><!--header-->
+<div class="contents">
+
+<p>This is the complete list of members for <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">tvm::runtime::metadata::Metadata</a>, including all inherited members.</p>
+<table class="directory">
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a3e9b0901b6e01257b060a45e159cc37e">_type_is_nullable</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a2d76fa1fb628ff276a284e61123589c5">as</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aa5c355fbb7d2f7402ee360dba8a52cdd">ContainerType</a> typedef</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#ac261cdb80487fb29ac42b28678f8cbef">data_</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">protected</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a17d8d5ad92691f9e18e3e0ae8ef69e4f">defined</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#acd04bb22a6861e9952c344ee8547411f">DowncastNoCheck</a>(ObjectRef ref)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">protected</span><span class="mlabel">static</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a22e5bb9d64dbc773bb9263b70882239e">FFIClearAfterMove</a>(ObjectRef *ref)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">protected</span><span class="mlabel">static</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aadbc0886ffa80162ff31eefd0431ba09">get</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#ae423057ecf93c18714d17f53cd1d318f">get_mutable</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">protected</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aed593996e4076632450de8fde776707c">GetDataPtr</a>(const ObjectRef &ref)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">protected</span><span class="mlabel">static</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html#a27735af93ebc35d49585dd7bba5ac099">Metadata</a>(const struct ::TVMMetadata *data)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">tvm::runtime::metadata::Metadata</a></td><td class="entry"><span class="mlabel">explicit</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aa07c1f6d66a438ea950637d13ed09471">ObjectRef</a>()=default</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a6a7dd7404edf1c26f8dbd9bd92d03a02">ObjectRef</a>(ObjectPtr< Object > data)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">explicit</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aa1bd13a7185cb4b2b6bdde49416e8aa4">operator!=</a>(const ObjectRef &other) const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a3deeeac5827a88f375b8c6ae1039c219">operator-></a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a4744bf4a1b48f202d41b51dc5e08e6ee">operator<</a>(const ObjectRef &other) const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#affdf1b8cdb36e140de7b3ad7064e4617">operator==</a>(const ObjectRef &other) const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#ae31a5b9f40781d60a2901994ead700e8">same_as</a>(const ObjectRef &other) const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html#a47ca34f745f4f50e989493f56f34d098">TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS</a>(Metadata, MetadataBase, MetadataNode)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">tvm::runtime::metadata::Metadata</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html#a98bb2f373db5ffafce12493806cc53a9">tvm::runtime::metadata::MetadataBase::TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS</a>(MetadataBase, ::tvm::runtime::ObjectRef, MetadataBaseNode)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">tvm::runtime::metadata::MetadataBase</a></td><td class="entry"></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a4e7cdb1574b93a59e784d70aa47b8da7">unique</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#a0ae0da21d247cd87ea94fe3777c4405e">use_count</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+</table></div><!-- contents -->
+<!-- start footer part -->
+<hr class="footer"/><address class="footer"><small>
+Generated by  <a href="http://www.doxygen.org/index.html">
+<img class="footer" src="doxygen.png" alt="doxygen"/>
+</a> 1.8.13
+</small></address>
+</body>
+</html>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata.html
similarity index 82%
copy from docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html
copy to docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata.html
index 90d346049..e837169f5 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1Metadata.html
@@ -5,7 +5,7 @@
<meta http-equiv="X-UA-Compatible" content="IE=9"/>
<meta name="generator" content="Doxygen 1.8.13"/>
<meta name="viewport" content="width=device-width, initial-scale=1"/>
-<title>tvm: tvm::runtime::metadata::MetadataBase Class Reference</title>
+<title>tvm: tvm::runtime::metadata::Metadata Class Reference</title>
<link href="tabs.css" rel="stylesheet" type="text/css"/>
<script type="text/javascript" src="jquery.js"></script>
<script type="text/javascript" src="dynsections.js"></script>
@@ -58,39 +58,41 @@ $(function() {
<div id="nav-path" class="navpath">
<ul>
-<li class="navelem"><a class="el" href="namespacetvm.html">tvm</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime.html">runtime</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">metadata</a></li><li class="navelem"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a></li> </ul>
+<li class="navelem"><a class="el" href="namespacetvm.html">tvm</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime.html">runtime</a></li><li class="navelem"><a class="el" href="namespacetvm_1_1runtime_1_1metadata.html">metadata</a></li><li class="navelem"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a></li> </ul>
</div>
</div><!-- top -->
<div class="header">
<div class="summary">
<a href="#pub-methods">Public Member Functions</a> |
-<a href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase-members.html">List of all members</a> </div>
+<a href="classtvm_1_1runtime_1_1metadata_1_1Metadata-members.html">List of all members</a> </div>
<div class="headertitle">
-<div class="title">tvm::runtime::metadata::MetadataBase Class Reference</div> </div>
+<div class="title">tvm::runtime::metadata::Metadata Class Reference</div> </div>
</div><!--header-->
<div class="contents">
-<p>Reference class for the common <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> class.
- <a href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html#details">More...</a></p>
-
-<p><code>#include <<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a>></code></p>
+<p><code>#include <<a class="el" href="metadata_8h_source.html">metadata.h</a>></code></p>
<div class="dynheader">
-Inheritance diagram for tvm::runtime::metadata::MetadataBase:</div>
+Inheritance diagram for tvm::runtime::metadata::Metadata:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1runtime_1_1metadata_1_1MetadataBase__inherit__graph.svg" width="202" height="647"><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="classtvm_1_1runtime_1_1metadata_1_1Metadata__inherit__graph.svg" width="202" height="647"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
<div class="dynheader">
-Collaboration diagram for tvm::runtime::metadata::MetadataBase:</div>
+Collaboration diagram for tvm::runtime::metadata::Metadata:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1runtime_1_1metadata_1_1MetadataBase__coll__graph.svg" width="202" height="766"><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="classtvm_1_1runtime_1_1metadata_1_1Metadata__coll__graph.svg" width="202" height="936"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
<table class="memberdecls">
<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-methods"></a>
Public Member Functions</h2></td></tr>
-<tr class="memitem:a98bb2f373db5ffafce12493806cc53a9"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html#a98bb2f373db5ffafce12493806cc53a9">TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS</a> (<a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a>, ::<a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a>, <a [...]
-<tr class="separator:a98bb2f373db5ffafce12493806cc53a9"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a27735af93ebc35d49585dd7bba5ac099"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html#a27735af93ebc35d49585dd7bba5ac099">Metadata</a> (const struct ::<a class="el" href="structTVMMetadata.html">TVMMetadata</a> *data)</td></tr>
+<tr class="separator:a27735af93ebc35d49585dd7bba5ac099"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a47ca34f745f4f50e989493f56f34d098"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html#a47ca34f745f4f50e989493f56f34d098">TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS</a> (<a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a>, <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a>, <a class="el [...]
+<tr class="separator:a47ca34f745f4f50e989493f56f34d098"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="inherit_header pub_methods_classtvm_1_1runtime_1_1metadata_1_1MetadataBase"><td colspan="2" onclick="javascript:toggleInherit('pub_methods_classtvm_1_1runtime_1_1metadata_1_1MetadataBase')"><img src="closed.png" alt="-"/> Public Member Functions inherited from <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">tvm::runtime::metadata::MetadataBase</a></td></tr>
+<tr class="memitem:a98bb2f373db5ffafce12493806cc53a9 inherit pub_methods_classtvm_1_1runtime_1_1metadata_1_1MetadataBase"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html#a98bb2f373db5ffafce12493806cc53a9">TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS</a> (<a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a>, ::<a class="el" href="cl [...]
+<tr class="separator:a98bb2f373db5ffafce12493806cc53a9 inherit pub_methods_classtvm_1_1runtime_1_1metadata_1_1MetadataBase"><td class="memSeparator" colspan="2"> </td></tr>
<tr class="inherit_header pub_methods_classtvm_1_1runtime_1_1ObjectRef"><td colspan="2" onclick="javascript:toggleInherit('pub_methods_classtvm_1_1runtime_1_1ObjectRef')"><img src="closed.png" alt="-"/> Public Member Functions inherited from <a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a></td></tr>
<tr class="memitem:aa07c1f6d66a438ea950637d13ed09471 inherit pub_methods_classtvm_1_1runtime_1_1ObjectRef"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#aa07c1f6d66a438ea950637d13ed09471">ObjectRef</a> ()=default</td></tr>
<tr class="memdesc:aa07c1f6d66a438ea950637d13ed09471 inherit pub_methods_classtvm_1_1runtime_1_1ObjectRef"><td class="mdescLeft"> </td><td class="mdescRight">default constructor <a href="classtvm_1_1runtime_1_1ObjectRef.html#aa07c1f6d66a438ea950637d13ed09471">More...</a><br /></td></tr>
@@ -154,31 +156,56 @@ Additional Inherited Members</h2></td></tr>
<tr class="memdesc:ac261cdb80487fb29ac42b28678f8cbef inherit pro_attribs_classtvm_1_1runtime_1_1ObjectRef"><td class="mdescLeft"> </td><td class="mdescRight">Internal pointer that backs the reference. <a href="classtvm_1_1runtime_1_1ObjectRef.html#ac261cdb80487fb29ac42b28678f8cbef">More...</a><br /></td></tr>
<tr class="separator:ac261cdb80487fb29ac42b28678f8cbef inherit pro_attribs_classtvm_1_1runtime_1_1ObjectRef"><td class="memSeparator" colspan="2"> </td></tr>
</table>
-<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
-<div class="textblock"><p>Reference class for the common <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> class. </p>
-</div><h2 class="groupheader">Member Function Documentation</h2>
-<a id="a98bb2f373db5ffafce12493806cc53a9"></a>
-<h2 class="memtitle"><span class="permalink"><a href="#a98bb2f373db5ffafce12493806cc53a9">◆ </a></span>TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS()</h2>
+<h2 class="groupheader">Constructor & Destructor Documentation</h2>
+<a id="a27735af93ebc35d49585dd7bba5ac099"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a27735af93ebc35d49585dd7bba5ac099">◆ </a></span>Metadata()</h2>
<div class="memitem">
<div class="memproto">
+<table class="mlabels">
+ <tr>
+ <td class="mlabels-left">
<table class="memname">
<tr>
- <td class="memname">tvm::runtime::metadata::MetadataBase::TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS </td>
+ <td class="memname">tvm::runtime::metadata::Metadata::Metadata </td>
<td>(</td>
- <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a> </td>
+ <td class="paramtype">const struct ::<a class="el" href="structTVMMetadata.html">TVMMetadata</a> * </td>
+ <td class="paramname"><em>data</em></td><td>)</td>
+ <td></td>
+ </tr>
+ </table>
+ </td>
+ <td class="mlabels-right">
+<span class="mlabels"><span class="mlabel">explicit</span></span> </td>
+ </tr>
+</table>
+</div><div class="memdoc">
+
+</div>
+</div>
+<h2 class="groupheader">Member Function Documentation</h2>
+<a id="a47ca34f745f4f50e989493f56f34d098"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a47ca34f745f4f50e989493f56f34d098">◆ </a></span>TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS()</h2>
+
+<div class="memitem">
+<div class="memproto">
+ <table class="memname">
+ <tr>
+ <td class="memname">tvm::runtime::metadata::Metadata::TVM_DEFINE_MUTABLE_OBJECT_REF_METHODS </td>
+ <td>(</td>
+ <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> </td>
<td class="paramname">, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
- <td class="paramtype">::<a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">tvm::runtime::ObjectRef</a> </td>
+ <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html">MetadataBase</a> </td>
<td class="paramname">, </td>
</tr>
<tr>
<td class="paramkey"></td>
<td></td>
- <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html">MetadataBaseNode</a> </td>
+ <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataNode.html">MetadataNode</a> </td>
<td class="paramname"> </td>
</tr>
<tr>
@@ -192,7 +219,7 @@ Additional Inherited Members</h2></td></tr>
</div>
</div>
<hr/>The documentation for this class was generated from the following file:<ul>
-<li>include/tvm/runtime/<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a></li>
+<li>include/tvm/runtime/<a class="el" href="metadata_8h_source.html">metadata.h</a></li>
</ul>
</div><!-- contents -->
<!-- start footer part -->
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html
index 1287e6b80..8ddce1cf0 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html
@@ -133,7 +133,7 @@ Public Attributes</h2></td></tr>
<tr class="memdesc:a695a21a69be1e72b330abe32c685552e"><td class="mdescLeft"> </td><td class="mdescRight">Describes the storage class of the emitted struct member. <a href="#a695a21a69be1e72b330abe32c685552e">More...</a><br /></td></tr>
<tr class="separator:a695a21a69be1e72b330abe32c685552e"><td class="memSeparator" colspan="2"> </td></tr>
<tr class="memitem:a452a726140b7e27bfdd647109f763be7"><td class="memItemLeft" align="right" valign="top">const char * </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html#a452a726140b7e27bfdd647109f763be7">type_key</a></td></tr>
-<tr class="memdesc:a452a726140b7e27bfdd647109f763be7"><td class="mdescLeft"> </td><td class="mdescRight">When <code>kind</code> is Metadata, type_key of the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> used with this array. <a href="#a452a726140b7e27bfdd647109f763be7">More...</a><br /></td></tr>
+<tr class="memdesc:a452a726140b7e27bfdd647109f763be7"><td class="mdescLeft"> </td><td class="mdescRight">When <code>kind</code> is <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a>, type_key of the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> used with this array. <a href="#a452a726140b7e27bfdd647109f763be7">More...</a><br /></td></tr>
<tr class="separator:a452a726140b7e27bfdd647109f763be7"><td class="memSeparator" colspan="2"> </td></tr>
</table><table class="memberdecls">
<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-static-attribs"></a>
@@ -397,7 +397,7 @@ Additional Inherited Members</h2></td></tr>
</table>
</div><div class="memdoc">
-<p>When <code>kind</code> is Metadata, type_key of the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> used with this array. </p>
+<p>When <code>kind</code> is <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a>, type_key of the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html" title="Common base class for all Metadata. ">MetadataBaseNode</a> used with this array. </p>
</div>
</div>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html
index 90d346049..137b7a645 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html
@@ -77,7 +77,7 @@ $(function() {
<div class="dynheader">
Inheritance diagram for tvm::runtime::metadata::MetadataBase:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1runtime_1_1metadata_1_1MetadataBase__inherit__graph.svg" width="202" height="647"><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="classtvm_1_1runtime_1_1metadata_1_1MetadataBase__inherit__graph.svg" width="631" height="647"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
<div class="dynheader">
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html
index 3018e7e36..a4bb8c235 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html
@@ -71,14 +71,14 @@ $(function() {
</div><!--header-->
<div class="contents">
-<p>Common base class for all Metadata.
+<p>Common base class for all <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a>.
<a href="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode.html#details">More...</a></p>
<p><code>#include <<a class="el" href="metadata__base_8h_source.html">metadata_base.h</a>></code></p>
<div class="dynheader">
Inheritance diagram for tvm::runtime::metadata::MetadataBaseNode:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg" width="287" height="926"><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="classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg" width="892" height="999"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe>
</div>
</div>
<div class="dynheader">
@@ -184,8 +184,8 @@ Additional Inherited Members</h2></td></tr>
<tr class="separator:af4407d2b59132e803ff791482dbe0145 inherit pro_attribs_classtvm_1_1runtime_1_1Object"><td class="memSeparator" colspan="2"> </td></tr>
</table>
<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
-<div class="textblock"><p>Common base class for all Metadata. </p>
-<p>This class is used in the visitor classes as a internal check to ensure that verify that all parts of the Metadata struct used in codegen are Metadata objects. </p>
+<div class="textblock"><p>Common base class for all <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a>. </p>
+<p>This class is used in the visitor classes as a internal check to ensure that verify that all parts of the <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> struct used in codegen are <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1Metadata.html">Metadata</a> objects. </p>
</div><h2 class="groupheader">Member Function Documentation</h2>
<a id="a6da4a7005f7b8dbf7df33def762b3bae"></a>
<h2 class="memtitle"><span class="permalink"><a href="#a6da4a7005f7b8dbf7df33def762b3bae">◆ </a></span>get_c_struct_name()</h2>
@@ -210,7 +210,7 @@ Additional Inherited Members</h2></td></tr>
</table>
</div><div class="memdoc">
-<p>Implemented in <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html#a3547ae61b4c04ca7b0af1f55d92f1953">tvm::runtime::metadata::MetadataArrayNode</a>.</p>
+<p>Implemented in <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html#a3547ae61b4c04ca7b0af1f55d92f1953">tvm::runtime::metadata::MetadataArrayNode</a>, <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1TensorInfoNode.html#a70b383ace832db8fef7900b79ca6c76b">tvm::runtime::metadata::TensorInfoNode</a>, and <a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataNode.html#a164b77727f2a899717fc33f0e30ebafe">tvm::runtime::metadata::MetadataNode</a>.</p>
</div>
</div>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__coll__graph.svg b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__coll__graph.svg
index 9486d0bb9..c39e95607 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__coll__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__coll__graph.svg
@@ -9,9 +9,9 @@
<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 509)">
<title>tvm::runtime::metadata::MetadataBaseNode</title>
<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-509 269,-509 269,4 -4,4"/>
-<!-- Node3 -->
+<!-- Node5 -->
<g id="node1" class="node">
-<title>Node3</title>
+<title>Node5</title>
<polygon fill="#bfbfbf" stroke="#000000" points="0,-.5 0,-79.5 207,-79.5 207,-.5 0,-.5"/>
<text text-anchor="start" x="8" y="-67.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
<text text-anchor="middle" x="103.5" y="-56.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataBaseNode</text>
@@ -21,9 +21,9 @@
<text text-anchor="start" x="8" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
<text text-anchor="start" x="8" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
</g>
-<!-- Node4 -->
+<!-- Node6 -->
<g id="node2" class="node">
-<title>Node4</title>
+<title>Node6</title>
<g id="a_node2"><a xlink:href="classtvm_1_1runtime_1_1Object.html" target="_top" xlink:title="base class of all object containers. ">
<polygon fill="#ffffff" stroke="#000000" points="12,-117.5 12,-504.5 195,-504.5 195,-117.5 12,-117.5"/>
<text text-anchor="middle" x="103.5" y="-492.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
@@ -64,15 +64,15 @@
</a>
</g>
</g>
-<!-- Node4->Node3 -->
+<!-- Node6->Node5 -->
<g id="edge1" class="edge">
-<title>Node4->Node3</title>
+<title>Node6->Node5</title>
<path fill="none" stroke="#191970" d="M103.5,-107.1434C103.5,-97.2018 103.5,-87.9253 103.5,-79.6313"/>
<polygon fill="none" stroke="#191970" points="100.0001,-107.295 103.5,-117.2951 107.0001,-107.2951 100.0001,-107.295"/>
</g>
-<!-- Node4->Node4 -->
+<!-- Node6->Node6 -->
<g id="edge2" class="edge">
-<title>Node4->Node4</title>
+<title>Node6->Node6</title>
<path fill="none" stroke="#404040" d="M195.3625,-344.9248C206.0482,-338.6637 213,-327.3555 213,-311 213,-300.0112 209.8618,-291.3007 204.5615,-284.8687"/>
<polygon fill="none" stroke="#404040" points="204.5184,-284.8322 197.3548,-284.0056 195.3625,-277.0752 202.5261,-277.9017 204.5184,-284.8322"/>
<text text-anchor="middle" x="239" y="-308.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #deleter_</text>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg
index 2708492dc..903781f2f 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1metadata_1_1MetadataBaseNode__inherit__graph.svg
@@ -4,98 +4,156 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: tvm::runtime::metadata::MetadataBaseNode Pages: 1 -->
-<svg width="215pt" height="694pt"
- viewBox="0.00 0.00 215.00 694.00" 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(4 690)">
+<svg width="669pt" height="749pt"
+ viewBox="0.00 0.00 669.00 749.00" 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(4 745)">
<title>tvm::runtime::metadata::MetadataBaseNode</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-690 211,-690 211,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-745 665,-745 665,4 -4,4"/>
<!-- Node0 -->
<g id="node1" class="node">
<title>Node0</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="0,-171.5 0,-250.5 207,-250.5 207,-171.5 0,-171.5"/>
-<text text-anchor="start" x="8" y="-238.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
-<text text-anchor="middle" x="103.5" y="-227.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataBaseNode</text>
-<polyline fill="none" stroke="#000000" points="0,-220.5 207,-220.5 "/>
-<text text-anchor="start" x="8" y="-208.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<polyline fill="none" stroke="#000000" points="0,-201.5 207,-201.5 "/>
-<text text-anchor="start" x="8" y="-189.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
-<text text-anchor="start" x="8" y="-178.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="226,-226.5 226,-305.5 433,-305.5 433,-226.5 226,-226.5"/>
+<text text-anchor="start" x="234" y="-293.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
+<text text-anchor="middle" x="329.5" y="-282.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataBaseNode</text>
+<polyline fill="none" stroke="#000000" points="226,-275.5 433,-275.5 "/>
+<text text-anchor="start" x="234" y="-263.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="226,-256.5 433,-256.5 "/>
+<text text-anchor="start" x="234" y="-244.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
+<text text-anchor="start" x="234" y="-233.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
</g>
<!-- Node2 -->
<g id="node3" class="node">
<title>Node2</title>
<g id="a_node3"><a xlink:href="classtvm_1_1runtime_1_1metadata_1_1MetadataArrayNode.html" target="_top" xlink:title="Container for arrays in the metadata. ">
-<polygon fill="#ffffff" stroke="#000000" points="0,-.5 0,-134.5 207,-134.5 207,-.5 0,-.5"/>
-<text text-anchor="start" x="8" y="-122.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
-<text text-anchor="middle" x="103.5" y="-111.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataArrayNode</text>
-<polyline fill="none" stroke="#000000" points="0,-104.5 207,-104.5 "/>
-<text text-anchor="start" x="8" y="-92.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ array</text>
-<text text-anchor="start" x="8" y="-81.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ kind</text>
-<text text-anchor="start" x="8" y="-70.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_key</text>
-<text text-anchor="start" x="8" y="-59.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<polyline fill="none" stroke="#000000" points="0,-52.5 207,-52.5 "/>
-<text text-anchor="start" x="8" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ MetadataArrayNode()</text>
-<text text-anchor="start" x="8" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
-<text text-anchor="start" x="8" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_element_c_struct_name()</text>
-<text text-anchor="start" x="8" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
+<polygon fill="#ffffff" stroke="#000000" points="0,-28 0,-162 207,-162 207,-28 0,-28"/>
+<text text-anchor="start" x="8" y="-150" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
+<text text-anchor="middle" x="103.5" y="-139" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataArrayNode</text>
+<polyline fill="none" stroke="#000000" points="0,-132 207,-132 "/>
+<text text-anchor="start" x="8" y="-120" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ array</text>
+<text text-anchor="start" x="8" y="-109" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ kind</text>
+<text text-anchor="start" x="8" y="-98" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_key</text>
+<text text-anchor="start" x="8" y="-87" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="0,-80 207,-80 "/>
+<text text-anchor="start" x="8" y="-68" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ MetadataArrayNode()</text>
+<text text-anchor="start" x="8" y="-57" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
+<text text-anchor="start" x="8" y="-46" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_element_c_struct_name()</text>
+<text text-anchor="start" x="8" y="-35" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
</a>
</g>
</g>
<!-- Node0->Node2 -->
<g id="edge2" class="edge">
<title>Node0->Node2</title>
-<path fill="none" stroke="#191970" d="M103.5,-161.173C103.5,-152.5614 103.5,-143.4836 103.5,-134.5054"/>
-<polygon fill="none" stroke="#191970" points="100.0001,-161.4081 103.5,-171.4081 107.0001,-161.4082 100.0001,-161.4081"/>
+<path fill="none" stroke="#191970" d="M258.8652,-220.7523C244.3086,-210.9173 229.2368,-200.3505 215.5,-190 203.8927,-181.2541 191.9018,-171.6931 180.2736,-162.1136"/>
+<polygon fill="none" stroke="#191970" points="257.1078,-223.788 267.3625,-226.4513 261.007,-217.9744 257.1078,-223.788"/>
+</g>
+<!-- Node3 -->
+<g id="node4" class="node">
+<title>Node3</title>
+<g id="a_node4"><a xlink:href="classtvm_1_1runtime_1_1metadata_1_1MetadataNode.html" target="_top" xlink:title="{tvm::runtime::metadata\l::MetadataNode\n|+ _type_key\l|+ MetadataNode()\l+ get_c_struct_name()\l+ version()\l+ num_inputs()\l+ inputs()\l+ num_outputs()\l+ outputs()\l+ num_pools()\l+ pools()\l+ mod_name()\l+ data()\l+ TVM_DECLARE_FINAL_OBJECT_INFO()\l}">
+<polygon fill="#ffffff" stroke="#000000" points="225,-.5 225,-189.5 434,-189.5 434,-.5 225,-.5"/>
+<text text-anchor="start" x="233" y="-177.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
+<text text-anchor="middle" x="329.5" y="-166.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::MetadataNode</text>
+<polyline fill="none" stroke="#000000" points="225,-159.5 434,-159.5 "/>
+<text text-anchor="start" x="233" y="-147.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="225,-140.5 434,-140.5 "/>
+<text text-anchor="start" x="233" y="-128.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ MetadataNode()</text>
+<text text-anchor="start" x="233" y="-117.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
+<text text-anchor="start" x="233" y="-106.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ version()</text>
+<text text-anchor="start" x="233" y="-95.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ num_inputs()</text>
+<text text-anchor="start" x="233" y="-84.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ inputs()</text>
+<text text-anchor="start" x="233" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ num_outputs()</text>
+<text text-anchor="start" x="233" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ outputs()</text>
+<text text-anchor="start" x="233" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ num_pools()</text>
+<text text-anchor="start" x="233" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ pools()</text>
+<text text-anchor="start" x="233" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ mod_name()</text>
+<text text-anchor="start" x="233" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ data()</text>
+<text text-anchor="start" x="233" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
+</a>
+</g>
+</g>
+<!-- Node0->Node3 -->
+<g id="edge3" class="edge">
+<title>Node0->Node3</title>
+<path fill="none" stroke="#191970" d="M329.5,-216.33C329.5,-207.868 329.5,-198.8344 329.5,-189.6585"/>
+<polygon fill="none" stroke="#191970" points="326.0001,-216.3393 329.5,-226.3394 333.0001,-216.3394 326.0001,-216.3393"/>
+</g>
+<!-- Node4 -->
+<g id="node5" class="node">
+<title>Node4</title>
+<g id="a_node5"><a xlink:href="classtvm_1_1runtime_1_1metadata_1_1TensorInfoNode.html" target="_top" xlink:title="{tvm::runtime::metadata\l::TensorInfoNode\n|+ _type_key\l|+ TensorInfoNode()\l+ get_c_struct_name()\l+ name()\l+ num_shape()\l+ shape()\l+ dtype()\l+ data()\l+ TVM_DECLARE_FINAL_OBJECT_INFO()\l}">
+<polygon fill="#ffffff" stroke="#000000" points="452,-22.5 452,-167.5 661,-167.5 661,-22.5 452,-22.5"/>
+<text text-anchor="start" x="460" y="-155.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::metadata</text>
+<text text-anchor="middle" x="556.5" y="-144.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::TensorInfoNode</text>
+<polyline fill="none" stroke="#000000" points="452,-137.5 661,-137.5 "/>
+<text text-anchor="start" x="460" y="-125.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="452,-118.5 661,-118.5 "/>
+<text text-anchor="start" x="460" y="-106.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TensorInfoNode()</text>
+<text text-anchor="start" x="460" y="-95.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get_c_struct_name()</text>
+<text text-anchor="start" x="460" y="-84.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ name()</text>
+<text text-anchor="start" x="460" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ num_shape()</text>
+<text text-anchor="start" x="460" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ shape()</text>
+<text text-anchor="start" x="460" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ dtype()</text>
+<text text-anchor="start" x="460" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ data()</text>
+<text text-anchor="start" x="460" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
+</a>
+</g>
+</g>
+<!-- Node0->Node4 -->
+<g id="edge4" class="edge">
+<title>Node0->Node4</title>
+<path fill="none" stroke="#191970" d="M400.0998,-220.7058C414.6589,-210.874 429.7406,-200.3204 443.5,-190 452.9315,-182.9258 462.6221,-175.3266 472.1741,-167.6194"/>
+<polygon fill="none" stroke="#191970" points="397.9583,-217.9277 391.6019,-226.404 401.8568,-223.7416 397.9583,-217.9277"/>
</g>
<!-- Node1 -->
<g id="node2" class="node">
<title>Node1</title>
<g id="a_node2"><a xlink:href="classtvm_1_1runtime_1_1Object.html" target="_top" xlink:title="base class of all object containers. ">
-<polygon fill="#ffffff" stroke="#000000" points="12,-287.5 12,-685.5 195,-685.5 195,-287.5 12,-287.5"/>
-<text text-anchor="middle" x="103.5" y="-673.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
-<polyline fill="none" stroke="#000000" points="12,-666.5 195,-666.5 "/>
-<text text-anchor="start" x="20" y="-654.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<text text-anchor="start" x="20" y="-643.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_final</text>
-<text text-anchor="start" x="20" y="-632.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots</text>
-<text text-anchor="start" x="20" y="-621.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots_can</text>
-<text text-anchor="start" x="20" y="-610.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_overflow</text>
-<text text-anchor="start" x="20" y="-599.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_visit</text>
-<text text-anchor="start" x="20" y="-588.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_attrs</text>
-<text text-anchor="start" x="20" y="-577.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_sequal</text>
-<text text-anchor="start" x="20" y="-566.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="20" y="-555.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_shash</text>
-<text text-anchor="start" x="20" y="-544.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="20" y="-533.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_index</text>
-<text text-anchor="start" x="20" y="-522.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># type_index_</text>
-<text text-anchor="start" x="20" y="-511.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># ref_counter_</text>
-<text text-anchor="start" x="20" y="-500.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># deleter_</text>
-<polyline fill="none" stroke="#000000" points="12,-493.5 195,-493.5 "/>
-<text text-anchor="start" x="20" y="-481.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_index()</text>
-<text text-anchor="start" x="20" y="-470.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKey()</text>
-<text text-anchor="start" x="20" y="-459.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKeyHash()</text>
-<text text-anchor="start" x="20" y="-448.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsInstance()</text>
-<text text-anchor="start" x="20" y="-437.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
-<text text-anchor="start" x="20" y="-426.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="20" y="-415.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="20" y="-404.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="20" y="-393.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="20" y="-382.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="20" y="-371.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2Key()</text>
-<text text-anchor="start" x="20" y="-360.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2KeyHash()</text>
-<text text-anchor="start" x="20" y="-349.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeKey2Index()</text>
-<text text-anchor="start" x="20" y="-338.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _GetOrAllocRuntimeTypeIndex()</text>
-<text text-anchor="start" x="20" y="-327.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RuntimeTypeIndex()</text>
... 41201 lines suppressed ...