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 2023/01/14 05:22:17 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@c452e6966c33047512155f42f63aef4e0586d129)
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 ea469b398b deploying docs (apache/tvm@c452e6966c33047512155f42f63aef4e0586d129)
ea469b398b is described below
commit ea469b398b178a10fdb1f320529f2068aece816e
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Sat Jan 14 05:22:10 2023 +0000
deploying docs (apache/tvm@c452e6966c33047512155f42f63aef4e0586d129)
---
docs/_images/sphx_glr_micro_train_001.png | Bin 327199 -> 337505 bytes
docs/_images/sphx_glr_micro_train_thumb.png | Bin 22934 -> 23675 bytes
.../how_to/compile_models/from_darknet.rst.txt | 2 +-
.../how_to/compile_models/from_keras.rst.txt | 2 +-
.../how_to/compile_models/from_mxnet.rst.txt | 2 +-
.../how_to/compile_models/from_oneflow.rst.txt | 2 +-
.../how_to/compile_models/from_pytorch.rst.txt | 2 +-
.../how_to/compile_models/from_tensorflow.rst.txt | 2 +-
.../compile_models/sg_execution_times.rst.txt | 22 +-
.../deploy_models/deploy_model_on_adreno.rst.txt | 2 +-
.../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 | 20 +-
.../extend_tvm/bring_your_own_datatypes.rst.txt | 2 +-
.../how_to/extend_tvm/sg_execution_times.rst.txt | 8 +-
.../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 | 14 +-
.../tune_conv2d_layer_cuda.rst.txt | 1389 +----
.../tune_network_cuda.rst.txt | 4 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 162 +-
.../tune_with_autotvm/sg_execution_times.rst.txt | 4 +-
.../tune_with_autotvm/tune_conv2d_cuda.rst.txt | 449 +-
.../work_with_microtvm/micro_autotune.rst.txt | 16 +-
.../work_with_microtvm/micro_pytorch.rst.txt | 4 +-
.../how_to/work_with_microtvm/micro_train.rst.txt | 18 +-
.../work_with_microtvm/sg_execution_times.rst.txt | 16 +-
.../work_with_relay/sg_execution_times.rst.txt | 8 +-
.../how_to/work_with_schedules/intrin_math.rst.txt | 2 +-
.../how_to/work_with_schedules/reduction.rst.txt | 2 +-
.../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 | 4 +-
.../frontend/deploy_classification.rst.txt | 2 +-
.../tutorials/frontend/deploy_detection.rst.txt | 2 +-
.../tutorials/frontend/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/matrix_multiply.rst.txt | 2 +-
.../tutorials/optimize/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/vta_get_started.rst.txt | 2 +-
.../tutorial/auto_scheduler_matmul_x86.rst.txt | 13 +-
docs/_sources/tutorial/autotvm_matmul_x86.rst.txt | 20 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 54 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 22 +-
.../tutorial/tensor_expr_get_started.rst.txt | 45 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_darknet.html | 2 +-
docs/how_to/compile_models/from_keras.html | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 13 +-
docs/how_to/compile_models/from_pytorch.html | 10 +-
docs/how_to/compile_models/from_tensorflow.html | 2 +-
docs/how_to/compile_models/sg_execution_times.html | 22 +-
.../deploy_models/deploy_model_on_adreno.html | 2 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 41 +-
docs/how_to/deploy_models/deploy_prequantized.html | 8 +-
.../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 | 37 +-
docs/how_to/deploy_models/sg_execution_times.html | 20 +-
.../extend_tvm/bring_your_own_datatypes.html | 2 +-
docs/how_to/extend_tvm/sg_execution_times.html | 8 +-
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 | 1389 +----
.../tune_with_autoscheduler/tune_network_cuda.html | 4 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 162 +-
.../tune_with_autotvm/sg_execution_times.html | 4 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.html | 449 +-
docs/how_to/work_with_microtvm/micro_autotune.html | 16 +-
docs/how_to/work_with_microtvm/micro_pytorch.html | 4 +-
docs/how_to/work_with_microtvm/micro_train.html | 16 +-
.../work_with_microtvm/sg_execution_times.html | 16 +-
.../how_to/work_with_relay/sg_execution_times.html | 8 +-
docs/how_to/work_with_schedules/intrin_math.html | 2 +-
docs/how_to/work_with_schedules/reduction.html | 2 +-
.../work_with_schedules/sg_execution_times.html | 18 +-
docs/how_to/work_with_schedules/tensorize.html | 2 +-
docs/reference/api/doxygen/classes.html | 28 +-
.../doxygen/classtvm_1_1runtime_1_1ObjectRef.html | 2 +-
...asstvm_1_1runtime_1_1ObjectRef__coll__graph.svg | 12 +-
...classtvm_1_1script_1_1printer_1_1FrameNode.html | 2 +-
...script_1_1printer_1_1FrameNode__coll__graph.svg | 466 +-
..._1script_1_1printer_1_1IRDocsifier-members.html | 2 +-
...asstvm_1_1script_1_1printer_1_1IRDocsifier.html | 27 +-
...ript_1_1printer_1_1IRDocsifierNode-members.html | 45 +-
...vm_1_1script_1_1printer_1_1IRDocsifierNode.html | 23 +-
..._1_1printer_1_1IRDocsifierNode__coll__graph.svg | 270 +-
...1printer_1_1IRDocsifierNode__inherit__graph.svg | 89 +-
.../api/doxygen/data__type_8h_source.html | 2 +-
docs/reference/api/doxygen/functions_func_i.html | 2 +-
docs/reference/api/doxygen/functions_func_l.html | 2 +-
docs/reference/api/doxygen/functions_func_p.html | 3 +
docs/reference/api/doxygen/functions_func_s.html | 2 +-
docs/reference/api/doxygen/functions_func_t.html | 4 +-
docs/reference/api/doxygen/functions_func_v.html | 39 +-
docs/reference/api/doxygen/functions_i.html | 4 +-
docs/reference/api/doxygen/functions_l.html | 4 +-
docs/reference/api/doxygen/functions_m.html | 2 +-
docs/reference/api/doxygen/functions_p.html | 5 +-
docs/reference/api/doxygen/functions_s.html | 4 +-
docs/reference/api/doxygen/functions_t.html | 8 +-
docs/reference/api/doxygen/functions_u.html | 2 +-
docs/reference/api/doxygen/functions_v.html | 44 +-
docs/reference/api/doxygen/functions_vars_i.html | 2 +-
docs/reference/api/doxygen/functions_vars_v.html | 3 +
docs/reference/api/doxygen/hierarchy.html | 823 ++-
docs/reference/api/doxygen/inherit_graph_111.svg | 32 +-
docs/reference/api/doxygen/inherit_graph_12.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_120.svg | 6293 ++++++++++----------
docs/reference/api/doxygen/inherit_graph_207.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_208.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_41.svg | 16 +-
docs/reference/api/doxygen/inherit_graph_45.svg | 8 +-
docs/reference/api/doxygen/inherit_graph_94.svg | 8 +-
docs/reference/api/doxygen/inherits.html | 2 +-
.../api/doxygen/ir__docsifier_8h_source.html | 28 +-
.../api/doxygen/namespacemembers_func_l.html | 13 +-
.../api/doxygen/namespacemembers_func_s.html | 2 +-
docs/reference/api/doxygen/namespacemembers_l.html | 15 +-
docs/reference/api/doxygen/namespacemembers_s.html | 2 +-
.../doxygen/namespacetvm_1_1script_1_1printer.html | 17 +-
.../reference/api/doxygen/namespacetvm_1_1tir.html | 23 +
docs/reference/api/doxygen/printer_8h.html | 9 +-
docs/reference/api/doxygen/printer_8h__incl.svg | 1046 ++--
docs/reference/api/doxygen/printer_8h_source.html | 23 +-
docs/reference/api/doxygen/search/all_10.js | 2 +-
docs/reference/api/doxygen/search/all_11.js | 5 +-
docs/reference/api/doxygen/search/all_13.js | 6 +-
docs/reference/api/doxygen/search/all_14.js | 12 +-
docs/reference/api/doxygen/search/all_15.js | 14 +-
docs/reference/api/doxygen/search/all_16.js | 4 +-
docs/reference/api/doxygen/search/all_17.js | 2 +
docs/reference/api/doxygen/search/all_18.js | 2 +-
docs/reference/api/doxygen/search/all_a.js | 4 +-
docs/reference/api/doxygen/search/all_d.js | 3 +-
docs/reference/api/doxygen/search/all_e.js | 7 +-
docs/reference/api/doxygen/search/classes_10.js | 2 +-
docs/reference/api/doxygen/search/classes_11.js | 8 +-
docs/reference/api/doxygen/search/classes_a.js | 1 -
docs/reference/api/doxygen/search/classes_d.js | 2 +-
docs/reference/api/doxygen/search/classes_f.js | 2 +-
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 | 4 +-
docs/reference/api/doxygen/search/functions_15.js | 2 +-
docs/reference/api/doxygen/search/functions_16.js | 1 +
docs/reference/api/doxygen/search/functions_17.js | 2 +-
docs/reference/api/doxygen/search/functions_9.js | 2 +-
docs/reference/api/doxygen/search/functions_c.js | 3 +-
docs/reference/api/doxygen/search/functions_d.js | 4 +-
docs/reference/api/doxygen/search/variables_14.js | 1 +
docs/reference/api/doxygen/search/variables_9.js | 2 +-
...vm_1_1script_1_1printer_1_1Default-members.html | 4 +
.../structtvm_1_1script_1_1printer_1_1Default.html | 95 +-
..._1script_1_1printer_1_1Default__coll__graph.svg | 89 +-
docs/reference/api/doxygen/tir_2expr_8h.html | 3 +
.../reference/api/doxygen/tir_2expr_8h_source.html | 5 +-
docs/reference/api/python/auto_scheduler.html | 4 +-
docs/reference/api/python/relay/index.html | 2 +-
.../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 | 4 +-
.../tutorials/frontend/deploy_classification.html | 2 +-
.../vta/tutorials/frontend/deploy_detection.html | 2 +-
.../vta/tutorials/frontend/sg_execution_times.html | 10 +-
docs/topic/vta/tutorials/matrix_multiply.html | 2 +-
.../vta/tutorials/optimize/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/vta_get_started.html | 2 +-
docs/tutorial/auto_scheduler_matmul_x86.html | 8 +-
docs/tutorial/autotvm_matmul_x86.html | 20 +-
docs/tutorial/autotvm_relay_x86.html | 270 +-
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 28 +-
docs/tutorial/tensor_expr_get_started.html | 41 +-
217 files changed, 6903 insertions(+), 8541 deletions(-)
diff --git a/docs/_images/sphx_glr_micro_train_001.png b/docs/_images/sphx_glr_micro_train_001.png
index 9acba7fd3b..cdece017f1 100644
Binary files a/docs/_images/sphx_glr_micro_train_001.png and b/docs/_images/sphx_glr_micro_train_001.png differ
diff --git a/docs/_images/sphx_glr_micro_train_thumb.png b/docs/_images/sphx_glr_micro_train_thumb.png
index fb0f49ab60..3ea3b2a601 100644
Binary files a/docs/_images/sphx_glr_micro_train_thumb.png and b/docs/_images/sphx_glr_micro_train_thumb.png differ
diff --git a/docs/_sources/how_to/compile_models/from_darknet.rst.txt b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
index 2d7c6767d1..319cf93ec0 100644
--- a/docs/_sources/how_to/compile_models/from_darknet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
@@ -318,7 +318,7 @@ The process is no different from other examples.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 8.187 seconds)
+ **Total running time of the script:** ( 1 minutes 15.348 seconds)
.. _sphx_glr_download_how_to_compile_models_from_darknet.py:
diff --git a/docs/_sources/how_to/compile_models/from_keras.rst.txt b/docs/_sources/how_to/compile_models/from_keras.rst.txt
index 6f33beb78b..ef1b2f5575 100644
--- a/docs/_sources/how_to/compile_models/from_keras.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_keras.rst.txt
@@ -232,7 +232,7 @@ Look up prediction top 1 index in 1000 class synset.
.. code-block:: none
Relay top-1 id: 285, class name: Egyptian cat
-
1/1 [==============================] - ETA: 0s
1/1 [==============================] - 1s 930ms/step
+
1/1 [==============================] - ETA: 0s
1/1 [==============================] - 1s 947ms/step
Keras top-1 id: 285, class name: Egyptian cat
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 15b1605954..63360cb1c1 100644
--- a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
@@ -116,7 +116,7 @@ In this section, we download a pretrained imagenet model and classify an image.
.. code-block:: none
- Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip3df8cdd9-4f05-4c71-a420-c4b44ef8ddb3 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipc7c73fb1-e93f-4182-85ca-fb0516cad9c0 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 d6d4b6e108..945c674f0b 100644
--- a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
@@ -121,7 +121,7 @@ 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]
15%|#5 | 6.33M/41.5M [00:00<00:00, 43.2MB/s]
25%|##5 | 10.5M/41.5M [00:00<00:01, 27.4MB/s]
41%|#### | 17.0M/41.5M [00:00<00:00, 39.9MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 37.2MB/s]
78%|#######8 | 32.5M/41.5M [00:00<00:00, 49.7MB/s]
92%|#########2| 38.3M/41.5M [00:00<00:00, 48.1MB/s]
100%|##########| 41.5M/41.5M [00:01<00:00, 43.3MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
19%|#9 | 7.99M/41.5M [00:00<00:00, 45.9MB/s]
39%|###8 | 16.0M/41.5M [00:00<00:00, 61.9MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 68.0MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 67.7MB/s]
96%|#########6| 40.0M/41.5M [00:00<00:00, 71.2MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 68.8MB/s]
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 db72b4d35a..cd0049eaf6 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -101,7 +101,7 @@ Load a pretrained PyTorch model
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
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]
18%|#7 | 7.99M/44.7M [00:00<00:00, 54.6MB/s]
36%|###5 | 16.0M/44.7M [00:00<00:00, 57.8MB/s]
54%|#####3 | 24.0M/44.7M [00:00<00:00, 59.2MB/s]
72%|#######1 | 32.0M/44.7M [00:00<00:00, 57.0MB/s]
88%|########8 | 39.3M/44.7M [00:00<00:00, 62.5MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 60.6MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
30%|### | 13.6M/44.7M [00:00<00:00, 141MB/s]
61%|###### | 27.0M/44.7M [00:00<00:00, 93.6MB/s]
90%|########9 | 40.0M/44.7M [00:00<00:00, 106MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 106MB/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 9edc411600..abe73ab9c3 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -424,7 +424,7 @@ Run the corresponding model on tensorflow
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 10.219 seconds)
+ **Total running time of the script:** ( 1 minutes 20.833 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 936a892c33..910e685307 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
@@ -5,26 +5,26 @@
Computation times
=================
-**05:37.346** total execution time for **how_to_compile_models** files:
+**06:20.885** total execution time for **how_to_compile_models** files:
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:10.219 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:20.833 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:08.187 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:15.348 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:45.442 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:53.068 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:30.946 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:35.786 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:28.392 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:30.700 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:25.999 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:30.611 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:25.325 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:26.729 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:23.244 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:24.782 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:17.145 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:20.421 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.446 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.607 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
index cc444d4f44..2a487f5384 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
@@ -727,7 +727,7 @@ well as provides information about the model's performance
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2751.7866 2751.6081 2754.8158 2749.1211 1.9819
+ 2755.3963 2755.1467 2757.3894 2753.0723 1.4283
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 0d8db3f7b9..34a2d9c174 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
@@ -437,7 +437,7 @@ Execute on TVM
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 17.1610 16.8567 21.1197 16.0662 1.4243
+ 16.1355 16.1322 16.2570 15.9628 0.0838
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 0f20917839..68388f7e5a 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
@@ -130,7 +130,7 @@ Load pre-trained maskrcnn from torchvision and do tracing
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=MaskRCNN_ResNet50_FPN_Weights.COCO_V1`. You can also use `weights=MaskRCNN_ResNet50_FPN_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
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]
8%|8 | 14.0M/170M [00:00<00:01, 147MB/s]
16%|#6 | 28.0M/170M [00:00<00:01, 76.8MB/s]
22%|##1 | 37.0M/170M [00:00<00:02, 63.3MB/s]
26%|##5 | 43.9M/170M [00:00<00:02, 64.8MB/s]
30%|##9 | 50.8M/170M [00:00<00:02, 61.0MB/s]
34%|###3 | 56.9M/170M [00:00<00:02, 58.2MB/s]
38%|###7 | 64.0M/170M [00:01<00:01, 59.4MB/s]
42%|####2 | 72.1M/170M [00:01<00:01, 65.9MB/s]
47%|####7 | 80.0M/170M [00:01<00:01, 59.9MB/s]
52%|#####1 | 88.0M/170M [00:01<00:01, 62.0MB/s]
57%|#####6 | 96.0M/170M [00:01<00:01, 62.6MB/s]
61%|######1 | 104M/170M [00:01<00:01, 60.8MB/s]
66%|######5 | 112M/170M [00:01<00:00, 64.4MB/s]
73%|#######3 | 125M/170M [00:01<00:00, 81.7MB/s]
79%|#######8 | 134M/170M [00:02<00:00, 84.4MB/s]
84%|########3 | 142M/170M [00:02<00:00, 79.7MB/s]
88%|########8 | 150M/170M [00:02<00:00, 73.9MB/s]
92%|#########2| 157M/170M [00:02<00:00, 68.8MB/s]
96%|#########6| 164M/170M [00:02<00:00, 60.4MB/s]
100%|#########9| 170M/170M [00:02<00:00, 57.8MB/s]
100%|##########| 170M/170M [00:02<00:00, 66.2MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
5%|4 | 7.99M/170M [00:00<00:02, 66.6MB/s]
9%|9 | 16.0M/170M [00:00<00:02, 69.8MB/s]
14%|#4 | 24.0M/170M [00:00<00:02, 67.7MB/s]
20%|## | 34.1M/170M [00:00<00:01, 81.0MB/s]
25%|##4 | 42.1M/170M [00:00<00:01, 78.8MB/s]
29%|##9 | 49.7M/170M [00:00<00:01, 73.7MB/s]
34%|###4 | 57.9M/170M [00:00<00:01, 77.3MB/s]
40%|#### | 68.4M/170M [00:00<00:01, 87.0MB/s]
45%|####5 | 76.9M/170M [00:01<00:01, 86.9MB/s]
52%|#####1 | 87.6M/170M [00:01<00:00, 94.6MB/s]
57%|#####6 | 96.8M/170M [00:01<00:01, 75.6MB/s]
64%|######4 | 109M/170M [00:01<00:00, 89.6MB/s]
71%|####### | 120M/170M [00:01<00:00, 92.7MB/s]
79%|#######8 | 134M/170M [00:01<00:00, 106MB/s]
85%|########4 | 144M/170M [00:01<00:00, 97.1MB/s]
91%|######### | 154M/170M [00:01<00:00, 96.9MB/s]
97%|#########7| 165M/170M [00:01<00:00, 102MB/s]
100%|##########| 170M/170M [00:02<00:00, 88.9MB/s]
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: 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)
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: 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').
@@ -299,7 +299,7 @@ Get boxes with score larger than 0.9
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 10.544 seconds)
+ **Total running time of the script:** ( 3 minutes 36.065 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 90fc30c81a..25cb69ef26 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
@@ -227,7 +227,7 @@ training. Other models require a full post training calibration.
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=MobileNet_V2_Weights.IMAGENET1K_V1`. You can also use `weights=MobileNet_V2_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
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]
59%|#####8 | 7.99M/13.6M [00:00<00:00, 75.0MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 84.3MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
59%|#####8 | 7.99M/13.6M [00:00<00:00, 64.7MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 91.5MB/s]
@@ -409,7 +409,7 @@ Here we give an example of how to measure performance of TVM compiled models.
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 90.4797 90.3710 94.1902 90.1895 0.4371
+ 90.3827 90.3425 90.7219 90.0761 0.1844
@@ -458,7 +458,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 5.071 seconds)
+ **Total running time of the script:** ( 1 minutes 14.229 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 4aaf68bc8f..a3d85b0142 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
@@ -423,7 +423,7 @@ 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)
- 119.4774 119.3662 124.0014 117.3927 1.1786
+ 121.0030 120.9543 125.1200 119.8180 0.7057
@@ -460,7 +460,7 @@ 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:** ( 2 minutes 29.124 seconds)
+ **Total running time of the script:** ( 2 minutes 33.566 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 5a20810087..2a25b74c12 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -257,7 +257,7 @@ We create a Relay VM to build and execute the model.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 50.139 seconds)
+ **Total running time of the script:** ( 1 minutes 30.231 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 0b9a8f682d..0165f9ce0b 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
@@ -170,7 +170,7 @@ 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]
4%|4 | 5877/132723 [00:00<00:02, 58765.71KB/s]
11%|#1 | 14645/132723 [00:00<00:01, 75770.04KB/s]
18%|#7 | 23350/132723 [00:00<00:01, 80918.32KB/s]
24%|##4 | 32022/132723 [00:00<00:01, 83200.24KB/s]
31%|### | 40841/132723 [00:00<00:01, 84990.39KB/s]
37%|###7 | 49629/132723 [00:00<00:00, 85970.32KB/s]
44%|####4 | 58440/132723 [00:00<00:00, 86665.60KB/s]
51%|##### | 67188/132723 [00:00<00:00, 86923.48KB/s]
57%|#####7 | 75987/132723 [00:00<00:00, 87255.07KB/s]
64%|######3 | 84900/132723 [00:01<00:00, 87830.65KB/s]
71%|####### | 93784/132723 [00:01<00:00, 88138.60KB/s]
77%|#######7 | 102598/132723 [00:01<00:00, 88111.64KB/s]
84%|########3 | 111483/132723 [00:01<00:00, 88333.37KB/s]
91%|######### | 120324/132723 [00:01<00:00, 88354.31KB/s]
97%|#########7| 129196/132723 [00:01<00:00, 88456.73KB/s]
100%|#######
###| 132723/132723 [00:01<00:00, 86088.15KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
3%|2 | 3802/132723 [00:00<00:03, 38007.17KB/s]
8%|7 | 10553/132723 [00:00<00:02, 55357.30KB/s]
13%|#3 | 17833/132723 [00:00<00:01, 63314.99KB/s]
19%|#9 | 25223/132723 [00:00<00:01, 67491.97KB/s]
25%|##4 | 32549/132723 [00:00<00:01, 69568.22KB/s]
30%|### | 39915/132723 [00:00<00:01, 70956.56KB/s]
36%|###5 | 47300/132723 [00:00<00:01, 71900.70KB/s]
41%|####1 | 54627/132723 [00:00<00:01, 72334.79KB/s]
47%|####6 | 61862/132723 [00:00<00:00, 72334.56KB/s]
52%|#####2 | 69197/132723 [00:01<00:00, 72640.60KB/s]
58%|#####7 | 76578/132723 [00:01<00:00, 72995.81KB/s]
63%|######3 | 83988/132723 [00:01<00:00, 73330.31KB/s]
69%|######8 | 91322/132723 [00:01<00:00, 73231.57KB/s]
74%|#######4 | 98646/132723 [00:01<00:00, 73201.85KB/s]
80%|#######9 | 105998/132723 [00:01<00:00, 73293.99KB/s]
85%|########5
| 113345/132723 [00:01<00:00, 73345.21KB/s]
91%|######### | 120680/132723 [00:01<00:00, 73279.05KB/s]
96%|#########6| 128008/132723 [00:01<00:00, 73209.82KB/s]
100%|##########| 132723/132723 [00:01<00:00, 71088.57KB/s]
@@ -246,7 +246,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 2.230 seconds)
+ **Total running time of the script:** ( 3 minutes 34.278 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 389f61a208..8dc6b319eb 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
@@ -5,26 +5,26 @@
Computation times
=================
-**13:55.486** total execution time for **how_to_deploy_models** files:
+**15:01.388** total execution time for **how_to_deploy_models** files:
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:10.544 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:36.065 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:02.230 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:34.278 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:29.124 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:33.566 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:50.139 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:30.231 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:05.071 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:14.229 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:53.556 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:56.417 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:34.648 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:40.938 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:25.234 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:28.076 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:24.933 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:27.582 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``) | 00:00.006 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
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 1c618e6b34..f2f5f70fc9 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
@@ -463,7 +463,7 @@ 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.zip01d1a502-6542-48df-a991-b0312e20263c from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zipaca19498-ab89-4cdf-8bb5-543eb74514c2 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 80fa7ba18b..94abdfcf94 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
@@ -5,14 +5,14 @@
Computation times
=================
-**00:46.908** total execution time for **how_to_extend_tvm** files:
+**00:53.834** total execution time for **how_to_extend_tvm** files:
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:43.561 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:50.047 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.337 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.694 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.004 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.086 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``) | 00:00.007 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
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 be86c53eb3..4d1cb2096b 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
@@ -220,10 +220,10 @@ profile the execution time of each passes.
.. code-block:: none
Printing results of timing profile...
- InferType: 7368us [7368us] (47.01%; 47.01%)
- FoldScaleAxis: 8304us [6us] (52.99%; 52.99%)
- FoldConstant: 8299us [1679us] (52.95%; 99.93%)
- InferType: 6620us [6620us] (42.24%; 79.77%)
+ InferType: 18502us [18502us] (48.95%; 48.95%)
+ FoldScaleAxis: 19294us [9us] (51.05%; 51.05%)
+ FoldConstant: 19285us [1711us] (51.02%; 99.95%)
+ InferType: 17574us [17574us] (46.50%; 91.13%)
@@ -262,10 +262,10 @@ Refer to following sections and :py:func:`tvm.instrument.pass_instrument` for th
.. code-block:: none
Printing results of timing profile...
- InferType: 6620us [6620us] (44.90%; 44.90%)
- FoldScaleAxis: 8124us [4us] (55.10%; 55.10%)
- FoldConstant: 8120us [1655us] (55.07%; 99.95%)
- InferType: 6465us [6465us] (43.84%; 79.62%)
+ InferType: 17911us [17911us] (47.93%; 47.93%)
+ FoldScaleAxis: 19456us [8us] (52.07%; 52.07%)
+ FoldConstant: 19448us [1782us] (52.05%; 99.96%)
+ InferType: 17667us [17667us] (47.28%; 90.84%)
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 25a8de954f..9a872e6109 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
@@ -331,7 +331,7 @@ latency of convolution.
.. code-block:: none
- Convolution: 54.238945 ms
+ Convolution: 42.309631 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 fa58b373fa..8e3d82d6d6 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
@@ -660,7 +660,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 8.092464 ms
+ conv2d with tensor core: 13.360538 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 37392d66ad..dd248c4e60 100644
--- a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
@@ -134,8 +134,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
.. code-block:: none
- Numpy running time: 0.018495
- Baseline: 3.193068
+ Numpy running time: 0.019325
+ Baseline: 3.239618
@@ -229,7 +229,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.300526
+ Opt1: 0.331288
@@ -331,7 +331,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.332444
+ Opt2: 0.344937
@@ -426,7 +426,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.113062
+ Opt3: 0.128089
@@ -550,7 +550,7 @@ flattening.
.. code-block:: none
- Opt4: 0.110256
+ Opt4: 0.110026
@@ -671,7 +671,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.114555
+ Opt5: 0.112031
@@ -795,7 +795,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
.. code-block:: none
- Opt6: 0.152187
+ Opt6: 0.146417
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 0af7932a1f..258bf685f9 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
@@ -5,12 +5,12 @@
Computation times
=================
-**00:34.650** total execution time for **how_to_optimize_operators** files:
+**00:35.210** total execution time for **how_to_optimize_operators** files:
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:31.892 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.575 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.578 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.552 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.180 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.083 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
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 b1ffe936b4..d68d2ba4f3 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
@@ -5,18 +5,18 @@
Computation times
=================
-**08:59.047** total execution time for **how_to_tune_with_autoscheduler** files:
+**09:27.857** total execution time for **how_to_tune_with_autoscheduler** files:
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:35.421 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:44.391 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:30.882 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:40.523 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:01.117 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:06.185 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:28.896 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:29.281 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:11.819 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.284 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:10.913 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.192 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
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 95a7c7f1d8..aed4990187 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
@@ -242,686 +242,51 @@ cooperative fetching, unrolling and operator fusion.
bias: Buffer(bias_2: Pointer(float32), float32, [1, 512, 1, 1], []),
compute: Buffer(compute_2: Pointer(float32), float32, [1, 512, 7, 7], [])}
buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
- attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 28;
- allocate(conv2d_nchw: Pointer(local float32), float32, [7]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [7], [], scope="local", align=16)[0] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 32;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [2]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [252]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [192]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [1], [], scope="local", align=4)[0] = 0f32
conv2d_nchw_1[1] = 0f32
- conv2d_nchw_1[2] = 0f32
- conv2d_nchw_1[3] = 0f32
- conv2d_nchw_1[4] = 0f32
- conv2d_nchw_1[5] = 0f32
- conv2d_nchw_1[6] = 0f32
- for (rc.outer.outer: int32, 0, 64) {
- let cse_var_1: int32 = (rc.outer.outer*72)
- {
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((1 <= floormod(blockIdx.x, 7)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
+ for (rc.outer.outer: int32, 0, 128) {
+ for (ry.outer.outer: int32, 0, 3) {
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392 {
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [252], [], scope="shared")[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1*4), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1*4), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1*4), 9))) && (floormod((threadIdx.x_1*4), 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((rc.outer.outer*196) + (floordiv((threadIdx.x_1*4), 9)*7)) + (ry.outer.outer* [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 1), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 1), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 1), 9))) && (floormod(((threadIdx.x_1*4) + 1), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 1), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1*4) + 1), 9)) - 8)], 0f32, dtype= [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 2), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 2), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 2), 9))) && (floormod(((threadIdx.x_1*4) + 2), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 2), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1*4) + 2), 9)) - 8)], 0f32, dtype= [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 3), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 3), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 3), 9))) && (floormod(((threadIdx.x_1*4) + 3), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 3), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1*4) + 3), 9)) - 8)], 0f32, dtype= [...]
+ }
}
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73728)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147456)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221184)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294912)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368640)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442368)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516096)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else(((1 <= floormod(threadIdx.x_1, 9)) && (floormod(threadIdx.x_1, 9) < 8)), data_3[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) - 1)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392;
+ if @tir.likely((threadIdx.x_2 < 192), dtype=bool) {
+ kernel.shared_1: Buffer(kernel.shared, float32, [192], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 12)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(threadIdx.x_2, 12), 3)*9)) + (ry.outer.outer*3)) + floormod(threadIdx.x_2, 3))]
}
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[threadIdx.x_2] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73731)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147459)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221187)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294915)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368643)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442371)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516099)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else((((floormod(blockIdx.x, 7) < 6) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) + 6)], 0f32, dtype=float32)
+ for (rc.outer.inner: int32, 0, 2) {
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7))]*kernel.shared_1[((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6))]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7))]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 96)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 3)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 99)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 1)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 97)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 64)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 4)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 64)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 100)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 2)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 98)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 65)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 5)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 65)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 101)]))
}
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[threadIdx.x_2] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73734)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147462)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221190)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294918)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368646)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442374)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516102)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
}
}
- for (i3.inner: int32, 0, 7) {
- compute_3: Buffer(compute_2, float32, [25088], [])[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[i3.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
- }
+ compute_3: Buffer(compute_2, float32, [25088], [])[((blockIdx.x*784) + threadIdx.x)] = max((conv2d_nchw_1[0] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*16) + floordiv(threadIdx.x, 49))]), 0f32)
+ compute_3[(((blockIdx.x*784) + threadIdx.x) + 392)] = max((conv2d_nchw_1[1] + bias_3[(((blockIdx.x*16) + floordiv(threadIdx.x, 49)) + 8)]), 0f32)
}
}
@@ -975,7 +340,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.322 ms
+ Execution time of this operator: 0.404 ms
@@ -1025,18 +390,18 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
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=1)
- conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=128)
- 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_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=8)
+ 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_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=7)
+ 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=2)
- conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, 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=1)
@@ -1046,13 +411,13 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
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=1)
- compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=128)
- compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+ compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=8)
+ 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_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=7)
- compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
+ compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
+ compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
compute_i3_o_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)
@@ -1072,14 +437,14 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
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)
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=128)
+ 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=392)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
- pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
- pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=128)
+ 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=392)
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", 1024)
+ s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 16)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -1097,613 +462,49 @@ 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__(128) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[7];
- __shared__ float pad_temp_shared[72];
- __shared__ float kernel_shared[3072];
+ extern "C" __global__ void __launch_bounds__(392) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[2];
+ __shared__ float pad_temp_shared[252];
+ __shared__ float kernel_shared[192];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = ((((1 <= (((int)blockIdx.x) % 7)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73728)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147456)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221184)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294912)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368640)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442368)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516096)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = (((1 <= (((int)threadIdx.x) % 9)) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 1)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73731)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147459)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221187)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294915)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368643)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442371)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516099)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = (((((((int)blockIdx.x) % 7) < 6) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) + 6)] : 0.000000e+00f);
+ for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
+ for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
+ __syncthreads();
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 <= ((((((int)threadIdx.x) * 4) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) * 4) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) * 4) % 9))) && (((((int)threadIdx.x) * 4) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) * 4) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 1) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 1) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 1) % 9))) && ((((((int)threadIdx.x) * 4) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 2) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 2) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 2) % 9))) && ((((((int)threadIdx.x) * 4) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 3) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 3) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 3) % 9))) && ((((((int)threadIdx.x) * 4) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 192) {
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 12) * 4608)) + (rc_outer_outer * 36)) + (((((int)threadIdx.x) % 12) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
+ }
+ __syncthreads();
+ for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7))] * kernel_shared[(((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6))]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7))] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 96)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 3)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 99)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 97)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 64)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 4)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 64)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 100)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 98)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 65)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 5)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 65)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 101)]));
+ }
}
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73734)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147462)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221190)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294918)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368646)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442374)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516102)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- }
- for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
- compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
}
+ compute[((((int)blockIdx.x) * 784) + ((int)threadIdx.x))] = max((conv2d_nchw[0] + bias[((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 49))]), 0.000000e+00f);
+ compute[(((((int)blockIdx.x) * 784) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[1] + bias[(((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 49)) + 8)]), 0.000000e+00f);
}
@@ -1764,7 +565,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 5 minutes 35.421 seconds)
+ **Total running time of the script:** ( 5 minutes 44.391 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 0cdaf062a7..17614211c7 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
@@ -647,7 +647,7 @@ 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)
- 7.9023 7.8957 7.9158 7.8954 0.0096
+ 7.8511 7.8500 7.8552 7.8482 0.0030
@@ -675,7 +675,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 1.117 seconds)
+ **Total running time of the script:** ( 1 minutes 6.185 seconds)
.. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_network_cuda.py:
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 c328fb2cc7..aabe3ac190 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
@@ -666,7 +666,7 @@ 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)
- 754.1883 754.1875 754.8836 753.4939 0.5674
+ 761.4761 760.0342 764.5453 759.8487 2.1716
@@ -694,7 +694,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 30.882 seconds)
+ **Total running time of the script:** ( 1 minutes 40.523 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 ec9b1be21e..fe9d886312 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
@@ -390,102 +390,78 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [128, 512], []),
compute: Buffer(compute_2: Pointer(float32), float32, [128, 512], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute} {
- for (i0.outer: int32, 0, 8) "parallel" {
- allocate(compute_3: Pointer(global float32), float32, [256]), storage_scope = global;
- for (i1.outer: int32, 0, 32) {
- for (i.inner.init: int32, 0, 16) {
- let cse_var_1: int32 = (i.inner.init*16)
- {
- compute_4: Buffer(compute_3, float32, [256], [])[cse_var_1] = 0f32
- compute_4[(cse_var_1 + 1)] = 0f32
- compute_4[(cse_var_1 + 2)] = 0f32
- compute_4[(cse_var_1 + 3)] = 0f32
- compute_4[(cse_var_1 + 4)] = 0f32
- compute_4[(cse_var_1 + 5)] = 0f32
- compute_4[(cse_var_1 + 6)] = 0f32
- compute_4[(cse_var_1 + 7)] = 0f32
- compute_4[(cse_var_1 + 8)] = 0f32
- compute_4[(cse_var_1 + 9)] = 0f32
- compute_4[(cse_var_1 + 10)] = 0f32
- compute_4[(cse_var_1 + 11)] = 0f32
- compute_4[(cse_var_1 + 12)] = 0f32
- compute_4[(cse_var_1 + 13)] = 0f32
- compute_4[(cse_var_1 + 14)] = 0f32
- compute_4[(cse_var_1 + 15)] = 0f32
- }
- }
- for (elem_idx: int32, 0, (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(i1.outer + 1)] - placeholder_15[i1.outer])) {
- for (i.inner: int32, 0, 16) {
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_2: int32 = (i.inner*16)
- compute_4[cse_var_2] = (compute_4[cse_var_2] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[i1.outer]*16) + (elem_idx*16))]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(((i0.outer*4096) + (i.inner*256)) + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_3: int32 = ((i.inner*16) + 1)
- compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 1)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_4: int32 = ((i.inner*16) + 2)
- compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 2)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_5: int32 = ((i.inner*16) + 3)
- compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 3)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_6: int32 = ((i.inner*16) + 4)
- compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 4)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_7: int32 = ((i.inner*16) + 5)
- compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 5)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_8: int32 = ((i.inner*16) + 6)
- compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 6)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_9: int32 = ((i.inner*16) + 7)
- compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 7)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_10: int32 = ((i.inner*16) + 8)
- compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 8)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_11: int32 = ((i.inner*16) + 9)
- compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 9)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_12: int32 = ((i.inner*16) + 10)
- compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 10)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_13: int32 = ((i.inner*16) + 11)
- compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 11)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_14: int32 = ((i.inner*16) + 12)
- compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 12)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_15: int32 = ((i.inner*16) + 13)
- compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 13)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_16: int32 = ((i.inner*16) + 14)
- compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 14)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
+ for (i0.outer.i1.outer.fused: int32, 0, 16) "parallel" {
+ allocate(compute_3: Pointer(global float32), float32, [4096]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 4) {
+ for (nb_j.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 32) {
+ let cse_var_1: int32 = (((i.outer.inner*1024) + (i.inner.init*32)) + (nb_j.inner*16))
+ {
+ compute_4: Buffer(compute_3, float32, [4096], [])[cse_var_1] = 0f32
+ compute_4[(cse_var_1 + 1)] = 0f32
+ compute_4[(cse_var_1 + 2)] = 0f32
+ compute_4[(cse_var_1 + 3)] = 0f32
+ compute_4[(cse_var_1 + 4)] = 0f32
+ compute_4[(cse_var_1 + 5)] = 0f32
+ compute_4[(cse_var_1 + 6)] = 0f32
+ compute_4[(cse_var_1 + 7)] = 0f32
+ compute_4[(cse_var_1 + 8)] = 0f32
+ compute_4[(cse_var_1 + 9)] = 0f32
+ compute_4[(cse_var_1 + 10)] = 0f32
+ compute_4[(cse_var_1 + 11)] = 0f32
+ compute_4[(cse_var_1 + 12)] = 0f32
+ compute_4[(cse_var_1 + 13)] = 0f32
+ compute_4[(cse_var_1 + 14)] = 0f32
+ compute_4[(cse_var_1 + 15)] = 0f32
+ }
}
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_17: int32 = ((i.inner*16) + 15)
- compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 15)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
+ for (elem_idx: int32, 0, let cse_var_2: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_2 + 1)] - placeholder_15[cse_var_2])) {
+ for (i.inner: int32, 0, 32) {
+ let cse_var_21: int32 = (elem_idx*16)
+ let cse_var_20: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner)
+ let cse_var_19: int32 = ((i.outer.inner*8192) + (i.inner*256))
+ let cse_var_18: int32 = (((i.outer.inner*1024) + (i.inner*32)) + (nb_j.inner*16))
+ let cse_var_17: int32 = (cse_var_18 + 9)
+ let cse_var_16: int32 = (cse_var_18 + 8)
+ let cse_var_15: int32 = (cse_var_18 + 7)
+ let cse_var_14: int32 = (cse_var_18 + 6)
+ let cse_var_13: int32 = (cse_var_18 + 5)
+ let cse_var_12: int32 = (cse_var_18 + 4)
+ let cse_var_11: int32 = (cse_var_18 + 3)
+ let cse_var_10: int32 = (cse_var_18 + 2)
+ let cse_var_9: int32 = (cse_var_18 + 15)
+ let cse_var_8: int32 = (cse_var_18 + 14)
+ let cse_var_7: int32 = (cse_var_18 + 13)
+ let cse_var_6: int32 = (cse_var_18 + 12)
+ let cse_var_5: int32 = (cse_var_18 + 11)
+ let cse_var_4: int32 = (cse_var_18 + 10)
+ let cse_var_3: int32 = (cse_var_18 + 1)
+ {
+ compute_4[cse_var_18] = (compute_4[cse_var_18] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[cse_var_20]*16) + cse_var_21)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(cse_var_19 + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 1)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 2)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 3)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 4)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 5)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 6)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 7)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 8)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 9)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 10)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 11)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 12)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 13)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 14)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 15)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ }
+ }
}
}
}
- for (i0.inner: int32, 0, 16) {
- for (i1.inner: int32, 0, 16) {
- let cse_var_18: int32 = ((((i0.outer*8192) + (i0.inner*512)) + (i1.outer*16)) + i1.inner)
- compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_18] = max((compute_4[((i0.inner*16) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_18]), 0f32)
+ for (i0.inner: int32, 0, 128) {
+ for (i1.inner: int32, 0, 32) {
+ let cse_var_22: int32 = (((i0.inner*512) + (i0.outer.i1.outer.fused*32)) + i1.inner)
+ compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_22] = max((compute_4[((i0.inner*32) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_22]), 0f32)
}
}
}
@@ -542,7 +518,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 2.200 ms
+ Execution time of this operator: 1.804 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 6912c0b09f..4bc397038b 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
=================
-**00:33.080** total execution time for **how_to_tune_with_autotvm** files:
+**00:35.819** total execution time for **how_to_tune_with_autotvm** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:33.046 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:35.785 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.021 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
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 6bc22c2ef7..07e41997a2 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
@@ -390,7 +390,7 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 256, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5246138
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,666420
No: 2 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -513,10 +513,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,837597
- No: 3 GFLOPS: 1.84/1.84 result: MeasureResult(costs=(0.1259390105,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.4146933555603027, timestamp=1673620329.442073) [('tile_f', [-1, 64, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1228550
- No: 4 GFLOPS: 28.01/28.01 result: MeasureResult(costs=(0.008264531615384615,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9725844860076904, timestamp=1673620331.1912718) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,493609
- No: 5 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9432354
+ No: 3 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -638,8 +636,9 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7478219
- No: 6 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3730260
+ No: 4 GFLOPS: 3.76/3.76 result: MeasureResult(costs=(0.06163596725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.509656190872192, timestamp=1673672256.9632695) [('tile_f', [-1, 1, 1, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4107784
+ No: 5 GFLOPS: 0.00/3.76 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -761,9 +760,9 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9627989
- No: 7 GFLOPS: 8.42/28.01 result: MeasureResult(costs=(0.02750208075,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4386982917785645, timestamp=1673620333.7890215) [('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3256671
- No: 8 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6703134
+ No: 6 GFLOPS: 86.57/86.57 result: MeasureResult(costs=(0.0026740413421052632,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.388312816619873, timestamp=1673672259.522572) [('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1393554
+ No: 7 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -885,9 +884,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 2, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9208051
- No: 9 GFLOPS: 2.10/28.01 result: MeasureResult(costs=(0.11018129725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.1899983882904053, timestamp=1673620336.2825646) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8546562
- No: 10 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2058261
+ No: 8 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1009,8 +1007,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10028778
- No: 11 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10227168
+ No: 9 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1132,9 +1130,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,407833
- No: 12 GFLOPS: 200.73/200.73 result: MeasureResult(costs=(0.0011533035517241378,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1821565628051758, timestamp=1673620337.0173085) [('tile_f', [-1, 2, 32, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3507991
- No: 13 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 8, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2476476
+ No: 10 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1256,8 +1253,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,375480
- No: 14 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 1, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1183142
+ No: 11 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1379,8 +1376,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 64, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3174867
- No: 15 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3651501
+ No: 12 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1502,8 +1499,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 4, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2115377
- No: 16 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7328306
+ No: 13 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1625,10 +1622,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 128, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7160658
- No: 17 GFLOPS: 239.04/239.04 result: MeasureResult(costs=(0.0009684793679245284,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2931921482086182, timestamp=1673620338.6904929) [('tile_f', [-1, 4, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3560522
- No: 18 GFLOPS: 0.95/239.04 result: MeasureResult(costs=(0.24484147625,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.071900367736816, timestamp=1673620342.299537) [('tile_f', [-1, 1, 1, 256]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,812236
- No: 19 GFLOPS: 0.00/239.04 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,495707
+ No: 14 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1750,8 +1745,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8625470
- No: 20 GFLOPS: 0.00/239.04 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 256, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9377333
+ No: 15 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1873,7 +1868,395 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 4, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4975596
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5716292
+ No: 16 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+ tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+ Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2222920
+ No: 17 GFLOPS: 0.00/86.57 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
+ return self.__get_result()
+ File "/usr/lib/python3.7/concurrent/futures/_base.py", line 384, in __get_result
+ raise self._exception
+ File "/usr/lib/python3.7/concurrent/futures/thread.py", line 57, in run
+ result = self.fn(*self.args, **self.kwargs)
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 432, in <lambda>
+ worker = lambda *args: self._worker_run(*args)
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 401, in _worker_run
+ return proc.recv()
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 309, in recv
+ raise TimeoutError()
+ TimeoutError
+
+ [('tile_f', [-1, 1, 2, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9078804
+ No: 18 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+ tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+ Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2920237
+ No: 19 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+ tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+ Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9207968
+ No: 20 GFLOPS: 77.94/86.57 result: MeasureResult(costs=(0.002970176054054054,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8064920902252197, timestamp=1673672270.9395628) [('tile_f', [-1, 2, 1, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3300321
@@ -1928,9 +2311,9 @@ and measure running time.
Finish loading 20 records
Best config:
- [('tile_f', [-1, 4, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3560522
+ [('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1393554
Finish loading 20 records
- Time cost of this operator: 0.001384
+ Time cost of this operator: 0.001601
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 9a50b87995..fe8ed3cd27 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
@@ -363,10 +363,10 @@ Timing the untuned program
########## Build without Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
- tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.0 98.724 (1, 2, 10, 10, 3) 2 1 [310.0]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.021 0.962 (1, 6, 10, 10) 1 1 [3.021]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.985 0.314 (1, 1, 10, 10, 3) 1 1 [0.985]
- Total_time - 314.007 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 312.3 98.731 (1, 2, 10, 10, 3) 2 1 [312.3]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.026 0.957 (1, 6, 10, 10) 1 1 [3.026]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.987 0.312 (1, 1, 10, 10, 3) 1 1 [0.987]
+ Total_time - 316.313 - - - - -
@@ -431,10 +431,10 @@ Timing the tuned program
########## Build with Autotuning ##########
Node Name Ops Time(us) Time(%) Shape Inputs Outputs Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
- tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 102.1 97.391 (1, 6, 10, 10, 1) 2 1 [102.1]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.775 1.693 (1, 6, 10, 10) 1 1 [1.775]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.96 0.916 (1, 1, 10, 10, 3) 1 1 [0.96]
- Total_time - 104.835 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 139.2 98.15 (1, 6, 10, 10, 1) 2 1 [139.2]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.781 1.256 (1, 6, 10, 10) 1 1 [1.781]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.842 0.594 (1, 3, 10, 10, 1) 1 1 [0.842]
+ Total_time - 141.823 - - - - -
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
index 755bc2ec9e..d97c0ee62e 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
@@ -117,7 +117,7 @@ download a cat image and preprocess it to use as the model input.
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/ao/quantization/utils.py:281: UserWarning: must run observer before calling calculate_qparams. Returning default values.
"must run observer before calling calculate_qparams. " +
Downloading: "https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
-
0%| | 0.00/3.42M [00:00<?, ?B/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 45.5MB/s]
+
0%| | 0.00/3.42M [00:00<?, ?B/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 45.6MB/s]
/workspace/python/tvm/relay/frontend/pytorch_utils.py:47: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
return LooseVersion(torch_ver) > ver
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/setuptools/_distutils/version.py:346: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
@@ -322,7 +322,7 @@ Look up prediction top 1 index in 1000 class synset.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 2.311 seconds)
+ **Total running time of the script:** ( 1 minutes 11.209 seconds)
.. _sphx_glr_download_how_to_work_with_microtvm_micro_pytorch.py:
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
index b48065b845..9d9edf58f5 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
@@ -218,7 +218,7 @@ take about **2 minutes** to download the Stanford Cars, while COCO 2017 validati
.. code-block:: none
- '/tmp/tmpbg21p4m5/images/random'
+ '/tmp/tmpv7f1wxl6/images/random'
@@ -309,7 +309,7 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. image-sg:: /how_to/work_with_microtvm/images/sphx_glr_micro_train_001.png
- :alt: [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0]
+ :alt: [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0]
:srcset: /how_to/work_with_microtvm/images/sphx_glr_micro_train_001.png
:class: sphx-glr-single-img
@@ -318,8 +318,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. code-block:: none
- /tmp/tmpbg21p4m5/images/target contains 8144 images
- /tmp/tmpbg21p4m5/images/random contains 5000 images
+ /tmp/tmpv7f1wxl6/images/target contains 8144 images
+ /tmp/tmpv7f1wxl6/images/random contains 5000 images
@@ -494,13 +494,13 @@ the time on our validation set).
.. code-block:: none
Epoch 1/3
- 328/328 - 47s - loss: 0.2198 - accuracy: 0.9223 - val_loss: 0.1192 - val_accuracy: 0.9600 - 47s/epoch - 143ms/step
+ 328/328 - 47s - loss: 0.2314 - accuracy: 0.9218 - val_loss: 0.1113 - val_accuracy: 0.9573 - 47s/epoch - 144ms/step
Epoch 2/3
- 328/328 - 43s - loss: 0.1058 - accuracy: 0.9609 - val_loss: 0.1799 - val_accuracy: 0.9407 - 43s/epoch - 132ms/step
+ 328/328 - 44s - loss: 0.1005 - accuracy: 0.9632 - val_loss: 0.1100 - val_accuracy: 0.9607 - 44s/epoch - 133ms/step
Epoch 3/3
- 328/328 - 43s - loss: 0.0606 - accuracy: 0.9779 - val_loss: 0.0981 - val_accuracy: 0.9687 - 43s/epoch - 132ms/step
+ 328/328 - 44s - loss: 0.0765 - accuracy: 0.9730 - val_loss: 0.1205 - val_accuracy: 0.9607 - 44s/epoch - 133ms/step
- <keras.callbacks.History object at 0x7fae16c89150>
+ <keras.callbacks.History object at 0x7f14b419db50>
@@ -857,7 +857,7 @@ Arduino tutorial for how to do that `on GitHub <https://github.com/guberti/tvm-a
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 4 minutes 53.524 seconds)
+ **Total running time of the script:** ( 4 minutes 28.535 seconds)
.. _sphx_glr_download_how_to_work_with_microtvm_micro_train.py:
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 599b6a2585..f03ca02f3c 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
@@ -5,22 +5,22 @@
Computation times
=================
-**06:58.377** total execution time for **how_to_work_with_microtvm** files:
+**06:46.352** total execution time for **how_to_work_with_microtvm** files:
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:53.524 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:28.535 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:02.311 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:11.209 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:50.938 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:53.735 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:07.791 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:08.938 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.812 | 0.0 MB |
-+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.py``) | 00:00.000 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.935 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.py``) | 00:00.000 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.py``) | 00:00.000 | 0.0 MB |
++---------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_microtvm_micro_reference_vm.py` (``micro_reference_vm.py``) | 00:00.000 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
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 5475bcc446..88262038d1 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
@@ -5,14 +5,14 @@
Computation times
=================
-**00:44.187** total execution time for **how_to_work_with_relay** files:
+**00:45.858** total execution time for **how_to_work_with_relay** files:
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.432 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:33.692 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.283 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.362 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.467 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.797 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``) | 00:00.006 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
index 31cb6a5e04..3ca727b31b 100644
--- a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
@@ -264,7 +264,7 @@ The following example customizes CUDA lowering rule for :code:`exp`.
.. code-block:: none
- <function my_cuda_math_rule at 0x7fae1663b950>
+ <function my_cuda_math_rule at 0x7f14b46ccb00>
diff --git a/docs/_sources/how_to/work_with_schedules/reduction.rst.txt b/docs/_sources/how_to/work_with_schedules/reduction.rst.txt
index f533213a16..8caafbcad8 100644
--- a/docs/_sources/how_to/work_with_schedules/reduction.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/reduction.rst.txt
@@ -316,7 +316,7 @@ the first axis of reduced result of B.f
.. code-block:: none
- [reduce(combiner=comm_reducer(result=[(x + y)], lhs=[x], rhs=[y], identity_element=[0f]), source=[B.rf[k.inner.v, ax0]], init=[], axis=[iter_var(k.inner.v, range(min=0, ext=16))], where=(bool)1, value_index=0)]
+ [T.reduce(T.comm_reducer(lambda x, y: x + y, [T.float32(0)]), source=[B.rf[k_inner_v, ax0]], init=[], axis=[T.iter_var(k_inner_v, T.Range(0, 16), "CommReduce", "")], condition=True, value_index=0)]
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 99e7deaa00..6277133b57 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
@@ -5,22 +5,22 @@
Computation times
=================
-**00:05.106** total execution time for **how_to_work_with_schedules** files:
+**00:07.535** total execution time for **how_to_work_with_schedules** files:
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:02.513 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:04.995 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.187 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.136 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.605 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.603 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.589 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.573 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.112 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.120 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.049 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.051 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.029 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.033 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``) | 00:00.023 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``) | 00:00.024 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
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 11f753ad74..3500b172e8 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -347,7 +347,7 @@ The importing needs to happen before the tensorized GEMV being executed.
B: Buffer(B_2: Pointer(float32), float32, [512, 64], []),
C: Buffer(C_2: Pointer(float32), float32, [1024, 512], [])}
buffer_map = {A_1: A, B_1: B, C_1: C} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpfqspl5_8/input0.cc'\nsource_filename = \"/tmp/tmpfqspl5_8/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/tmpls10z2ft/input0.cc'\nsource_filename = \"/tmp/tmpls10z2ft/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 40283b39f1..baa5e42f28 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
@@ -5,10 +5,10 @@
Computation times
=================
-**00:25.819** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:30.755** total execution time for **topic_vta_tutorials_autotvm** files:
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:25.812 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:30.748 | 0.0 MB |
+---------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``) | 00:00.007 | 0.0 MB |
+---------------------------------------------------------------------------------------+-----------+--------+
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 0c72adf160..6541b49741 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -293,7 +293,7 @@ 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 27.37s!
+ resnet18_v1 inference graph built in 33.01s!
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 b9bba5ba79..94455793d5 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -337,7 +337,7 @@ The compilation steps are:
/workspace/python/tvm/relay/build_module.py:348: DeprecationWarning: Please use input parameter mod (tvm.IRModule) instead of deprecated parameter mod (tvm.relay.function.Function)
DeprecationWarning,
- yolov3-tiny inference graph built in 19.08s!
+ yolov3-tiny inference graph built in 22.44s!
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 845f4daa3e..828602e6c5 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
@@ -5,10 +5,10 @@
Computation times
=================
-**01:30.105** total execution time for **topic_vta_tutorials_frontend** files:
+**01:39.274** total execution time for **topic_vta_tutorials_frontend** files:
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:45.881 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:49.919 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:44.224 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.355 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/matrix_multiply.rst.txt b/docs/_sources/topic/vta/tutorials/matrix_multiply.rst.txt
index 9979cd5032..7f11482d1f 100644
--- a/docs/_sources/topic/vta/tutorials/matrix_multiply.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/matrix_multiply.rst.txt
@@ -464,7 +464,7 @@ memory.
.. code-block:: none
- stage(C_buf, compute(C_buf, body=[reduce(combiner=comm_reducer(result=[(x + y)], lhs=[x], rhs=[y], identity_element=[0]), source=[(int32(A_buf[bo, ko, bi, ki])*int32(B_buf[co, ko, ci, ki]))], init=[], axis=[iter_var(ko, range(min=0, ext=16)), iter_var(ki, range(min=0, ext=16))], where=(bool)1, value_index=0)], axis=[iter_var(bo, range(min=0, ext=1)), iter_var(co, range(min=0, ext=16)), iter_var(bi, range(min=0, ext=1)), iter_var(ci, range(min=0, ext=16))], reduce_axis=[iter_var(ko, r [...]
+ stage(C_buf, compute(C_buf, body=[T.reduce(T.comm_reducer(lambda x, y: x + y, [0]), source=[T.Cast("int32", A_buf[bo, ko, bi, ki]) * T.Cast("int32", B_buf[co, ko, ci, ki])], init=[], axis=[T.iter_var(ko, T.Range(0, 16), "CommReduce", ""), T.iter_var(ki, T.Range(0, 16), "CommReduce", "")], condition=True, value_index=0)], axis=[T.iter_var(bo, T.Range(0, 1), "DataPar", ""), T.iter_var(co, T.Range(0, 16), "DataPar", ""), T.iter_var(bi, T.Range(0, 1), "DataPar", ""), T.iter_var(ci, T.Ran [...]
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 d9830295fa..444558a3b3 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
@@ -5,10 +5,10 @@
Computation times
=================
-**00:03.285** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.186** total execution time for **topic_vta_tutorials_optimize** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.789 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.706 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.496 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.479 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
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 8ad5f28a36..7f7b477ea5 100644
--- a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
Computation times
=================
-**00:00.904** total execution time for **topic_vta_tutorials** files:
+**00:00.839** total execution time for **topic_vta_tutorials** files:
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.488 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.445 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.416 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.394 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/vta_get_started.rst.txt b/docs/_sources/topic/vta/tutorials/vta_get_started.rst.txt
index ed02a230fb..fb2e99b084 100644
--- a/docs/_sources/topic/vta/tutorials/vta_get_started.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/vta_get_started.rst.txt
@@ -433,7 +433,7 @@ Set the intermediate tensors' scope to VTA's on-chip accumulator buffer
.. code-block:: none
- stage(C_buf, compute(C_buf, body=[(A_buf[i0, i1, i2, i3] + B_buf[i0, i1, i2, i3])], axis=[iter_var(i0, range(min=0, ext=1)), iter_var(i1, range(min=0, ext=64)), iter_var(i2, range(min=0, ext=1)), iter_var(i3, range(min=0, ext=16))], reduce_axis=[], tag=, attrs={}))
+ stage(C_buf, compute(C_buf, body=[A_buf[i0, i1, i2, i3] + B_buf[i0, i1, i2, i3]], axis=[T.iter_var(i0, T.Range(0, 1), "DataPar", ""), T.iter_var(i1, T.Range(0, 64), "DataPar", ""), T.iter_var(i2, T.Range(0, 1), "DataPar", ""), T.iter_var(i3, T.Range(0, 16), "DataPar", "")], reduce_axis=[], tag=, attrs={}))
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index 3d2ed62fcf..34fa46ef7c 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -207,6 +207,13 @@ trials, we can load the best schedule from the log file and apply it.
+.. rst-class:: sphx-glr-script-out
+
+ .. code-block:: none
+
+
+ *E
+
@@ -329,7 +336,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 94.923 ms
+ Execution time of this operator: 96.894 ms
@@ -429,7 +436,7 @@ resume the status and do more 5 trials.
Resume search:
/venv/apache-tvm-py3.7/lib/python3.7/site-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)
-
+ .T
@@ -447,7 +454,7 @@ operations.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 14.772 seconds)
+ **Total running time of the script:** ( 1 minutes 40.995 seconds)
.. _sphx_glr_download_tutorial_auto_scheduler_matmul_x86.py:
diff --git a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
index ad7e06612c..e9d59ca183 100644
--- a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
@@ -454,16 +454,16 @@ reduce variance, we take 5 measurements and average them.
waiting for device...
device available
Get devices for measurement successfully!
- No: 1 GFLOPS: 0.50/0.50 result: MeasureResult(costs=(0.536838092,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.787323951721191, timestamp=1673618926.7616878) [('tile_y', [-1, 32]), ('tile_x', [-1, 1])],None,5
- No: 2 GFLOPS: 11.50/11.50 result: MeasureResult(costs=(0.023335867200000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6244955062866211, timestamp=1673618927.3820775) [('tile_y', [-1, 128]), ('tile_x', [-1, 32])],None,57
- No: 3 GFLOPS: 1.55/11.50 result: MeasureResult(costs=(0.1730271418,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.997842311859131, timestamp=1673618931.1427608) [('tile_y', [-1, 32]), ('tile_x', [-1, 4])],None,25
- No: 4 GFLOPS: 12.69/12.69 result: MeasureResult(costs=(0.0211559824,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5739216804504395, timestamp=1673618932.4802542) [('tile_y', [-1, 256]), ('tile_x', [-1, 512])],None,98
- No: 5 GFLOPS: 12.58/12.69 result: MeasureResult(costs=(0.0213331896,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7440025806427002, timestamp=1673618934.1530273) [('tile_y', [-1, 64]), ('tile_x', [-1, 128])],None,76
- No: 6 GFLOPS: 3.93/12.69 result: MeasureResult(costs=(0.0682771566,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.389047384262085, timestamp=1673618935.4969237) [('tile_y', [-1, 64]), ('tile_x', [-1, 16])],None,46
- No: 7 GFLOPS: 8.74/12.69 result: MeasureResult(costs=(0.0307298558,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.9488058090209961, timestamp=1673618936.2442718) [('tile_y', [-1, 16]), ('tile_x', [-1, 64])],None,64
- No: 8 GFLOPS: 9.56/12.69 result: MeasureResult(costs=(0.028072618,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.8177468776702881, timestamp=1673618936.9497247) [('tile_y', [-1, 8]), ('tile_x', [-1, 128])],None,73
- No: 9 GFLOPS: 10.94/12.69 result: MeasureResult(costs=(0.024539916199999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5884792804718018, timestamp=1673618937.6528091) [('tile_y', [-1, 512]), ('tile_x', [-1, 512])],None,99
- No: 10 GFLOPS: 1.21/12.69 result: MeasureResult(costs=(0.22114467380000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.736809253692627, timestamp=1673618941.4462543) [('tile_y', [-1, 1]), ('tile_x', [-1, 2])],None,10
+ No: 1 GFLOPS: 9.86/9.86 result: MeasureResult(costs=(0.0272303968,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6965093612670898, timestamp=1673670693.8256564) [('tile_y', [-1, 8]), ('tile_x', [-1, 32])],None,53
+ No: 2 GFLOPS: 11.28/11.28 result: MeasureResult(costs=(0.023793215799999996,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6471128463745117, timestamp=1673670695.2542806) [('tile_y', [-1, 128]), ('tile_x', [-1, 32])],None,57
+ No: 3 GFLOPS: 3.65/11.28 result: MeasureResult(costs=(0.073611713,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4325134754180908, timestamp=1673670696.6854718) [('tile_y', [-1, 128]), ('tile_x', [-1, 16])],None,47
+ No: 4 GFLOPS: 2.96/11.28 result: MeasureResult(costs=(0.0906588482,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6873035430908203, timestamp=1673670699.18775) [('tile_y', [-1, 4]), ('tile_x', [-1, 8])],None,32
+ No: 5 GFLOPS: 0.50/11.28 result: MeasureResult(costs=(0.5346576624,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.78984546661377, timestamp=1673670708.1348746) [('tile_y', [-1, 32]), ('tile_x', [-1, 1])],None,5
+ No: 6 GFLOPS: 3.28/11.28 result: MeasureResult(costs=(0.0819252484,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.5582821369171143, timestamp=1673670709.6975408) [('tile_y', [-1, 32]), ('tile_x', [-1, 8])],None,35
+ No: 7 GFLOPS: 10.30/11.28 result: MeasureResult(costs=(0.026063373,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7062933444976807, timestamp=1673670711.176278) [('tile_y', [-1, 512]), ('tile_x', [-1, 128])],None,79
+ No: 8 GFLOPS: 11.77/11.77 result: MeasureResult(costs=(0.0228116384,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5946991443634033, timestamp=1673670711.7940698) [('tile_y', [-1, 256]), ('tile_x', [-1, 512])],None,98
+ No: 9 GFLOPS: 0.49/11.77 result: MeasureResult(costs=(0.5475377647999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.966180801391602, timestamp=1673670720.8952181) [('tile_y', [-1, 512]), ('tile_x', [-1, 1])],None,9
+ No: 10 GFLOPS: 3.00/11.77 result: MeasureResult(costs=(0.0895305652,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6567583084106445, timestamp=1673670722.5795941) [('tile_y', [-1, 2]), ('tile_x', [-1, 16])],None,41
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index 6701aec50c..b353a8050d 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -311,7 +311,7 @@ standard deviation.
.. code-block:: none
- {'mean': 520.7501513800526, 'median': 520.3653747001226, 'std': 2.5685746603934763}
+ {'mean': 523.5642570399989, 'median': 524.7381491499993, 'std': 3.304938946858366}
@@ -545,30 +545,30 @@ the tuning data to.
.. code-block:: none
-
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 12.98/ 16.09 GFLOPS | Progress: (4/20) | 8.24 s
[Task 1/25] Current/Best: 8.79/ 16.09 GFLOPS | Progress: (8/20) | 12.46 s
[Task 1/25] Current/Best: 13.08/ 16.09 GFLOPS | Progress: (12/20) | 15.79 s
[Task 1/25] Current/Best: 18.45/ 18.45 GFLOPS | Progress: (16/20) | 18.05 s
[Task 1/25] Current/Best: 16.88/ 18.45 GFLOPS | Progress: (20/20) | 20.91 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 8.21/ 18.68 GFLOPS | Progress: (4/20) | 3.13 s
[Task 2/25] Current/Best: 12.14/ 18.68 GFLOPS | Progress: (8/20) | 6.20 s
[Task 2/25] Current/Best: 6.00/ 18.68 GFLOPS | Progress: (12/20) | 7.71 s
[Task 2/25] Current/Best: 6.41/ 19.80 GFLOPS | Progress: (16/20) | 9.28 s
[Task 2/25] Current/Best: 10.86/ 19.80 GFLOPS | Progress: (20/20) | 11.12 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 10.13/ 12.36 GFLOPS | Progress: (4/20) | 4.01 s
[Task 3/25] Current/Best: 3.02/ 14.60 GFLOPS | Progress: (8/20) | 6.93 s
[Task 3/25] Current/Best: 17.47/ 17.47 GFLOPS | Progress: (12/20) | 9.33 s
[Task 3/25] Current/Best: 8.46/ 19.46 GFLOPS | Progress: (16/20) | 11.84 s
[Task 3/25] Current/Best: 13.63/ 19.46 GFLOPS | Progress: (20/20) | 14.17 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 12.44/ 13.88 GFLOPS | Progress: (4/20) | 4.82 s
[Task 4/25] Current/Best: 12.07/ 17.85 GFLOPS | Progress: (8/20) | 7.10 s
[Task 4/25] Current/Best: 10.91/ 17.85 GFLOPS | Progress: (12/20) | 15.92 s
[Task 4/25] Current/Best: 10.65/ 17.85 GFLOPS | Progress: (16/20) | 19.45 s
[Task 4/25] Current/Best: 12.98/ 19.98 GFLOPS | Progress: (20/20) | 22.66 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 16.82/ 16.82 GFLOPS | Progress: (4/20) | 3.72 s
[Task 5/25] Current/Best: 14.79/ 16.82 GFLOPS | Progress: (8/20) | 7.15 s
[Task 5/25] Current/Best: 13.12/ 16.82 GFLOPS | Progress: (12/20) | 9.79 s
[Task 5/25] Current/Best: 9.63/ 19.07 GFLOPS | Progress: (16/20) | 11.83 s
[Task 5/25] Current/Best: 1.58/ 19.07 GFLOPS | Progress: (20/20) | 15.55 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 9.44/ 15.26 GFLOPS | Progress: (4/20) | 4.38 s
[Task 6/25] Current/Best: 5.85/ 15.26 GFLOPS | Progress: (8/20) | 7.63 s
[Task 6/25] Current/Best: 21.39/ 21.39 GFLOPS | Progress: (12/20) | 10.95 s
[Task 6/25] Current/Best: 8.01/ 21.39 GFLOPS | Progress: (16/20) | 14.57 s
[Task 6/25] Current/Best: 5.92/ 21.39 GFLOPS | Progress: (20/20) | 17.12 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 12.29/ 12.29 GFLOPS | Progress: (4/20) | 4.38 s
[Task 7/25] Current/Best: 15.81/ 15.81 GFLOPS | Progress: (8/20) | 7.04 s
[Task 7/25] Current/Best: 6.01/ 15.81 GFLOPS | Progress: (12/20) | 9.98 s
[Task 7/25] Current/Best: 18.09/ 20.26 GFLOPS | Progress: (16/20) | 12.39 s
[Task 7/25] Current/Best: 15.09/ 20.26 GFLOPS | Progress: (20/20) | 14.98 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 12.94/ 13.60 GFLOPS | Progress: (4/20) | 4.10 s
[Task 8/25] Current/Best: 12.69/ 14.35 GFLOPS | Progress: (8/20) | 6.99 s
[Task 8/25] Current/Best: 9.32/ 14.35 GFLOPS | Progress: (12/20) | 15.93 s
[Task 8/25] Current/Best: 7.84/ 14.35 GFLOPS | Progress: (16/20) | 19.55 s
[Task 8/25] Current/Best: 9.85/ 14.35 GFLOPS | Progress: (20/20) | 23.62 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 13.71/ 19.65 GFLOPS | Progress: (4/20) | 4.81 s
[Task 9/25] Current/Best: 9.83/ 19.65 GFLOPS | Progress: (8/20) | 6.85 s
[Task 9/25] Current/Best: 18.01/ 19.65 GFLOPS | Progress: (12/20) | 11.51 s
[Task 9/25] Current/Best: 10.35/ 19.65 GFLOPS | Progress: (16/20) | 22.66 s
[Task 9/25] Current/Best: 5.86/ 19.65 GFLOPS | Progress: (20/20) | 25.21 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 4.08/ 18.23 GFLOPS | Progress: (4/20) | 3.65 s
[Task 10/25] Current/Best: 11.42/ 18.23 GFLOPS | Progress: (8/20) | 5.31 s
[Task 10/25] Current/Best: 20.97/ 20.97 GFLOPS | Progress: (12/20) | 6.92 s
[Task 10/25] Current/Best: 14.56/ 20.97 GFLOPS | Progress: (16/20) | 10.33 s
[Task 10/25] Current/Best: 10.24/ 20.97 GFLOPS | Progress: (20/20)
| 11.93 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 8.21/ 22.33 GFLOPS | Progress: (4/20) | 4.13 s
[Task 11/25] Current/Best: 17.77/ 22.33 GFLOPS | Progress: (8/20) | 6.70 s
[Task 11/25] Current/Best: 21.85/ 22.33 GFLOPS | Progress: (12/20) | 8.65 s
[Task 11/25] Current/Best: 12.58/ 22.33 GFLOPS | Progress: (16/20) | 12.66 s
[Task 11/25] Current/Best: 17.78/ 22.33 GFLOPS | Progress: (20/20) | 15.49 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 16.23/ 16.23 GFLOPS | Progress: (4/20) | 7.01 s
[Task 12/25] Current/Best: 10.25/ 20.88 GFLOPS | Progress: (8/20) | 9.69 s
[Task 12/25] Current/Best: 9.90/ 20.88 GFLOPS | Progress: (12/20) | 12.32 s
[Task 12/25] Current/Best: 18.80/ 20.88 GFLOPS | Progress: (16/20) | 14.51 s
[Task 12/25] Current/Best: 19.49/ 20.88 GFLOPS | Progress: (20/20) | 17.71 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 20.93/ 20.93 GFLOPS | Progress: (4/20) | 4.78 s
[Task 13/25] Current/Best: 18.42/ 20.93 GFLOPS | Progress: (8/20) | 7.90 s
[Task 13/25] Current/Best: 7.99/ 20.93 GFLOPS | Progress: (12/20) | 11.18 s
[Task 13/25] Current/Best: 9.89/ 20.93 GFLOPS | Progress: (16/20) | 14.19 s
[Task 13/25] Current/Best: 12.72/ 20.99 GFLOPS | Progress: (20/20) | 16.85 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 5.39/ 17.36 GFLOPS | Progress: (4/20) | 5.70 s
[Task 14/25] Current/Best: 20.04/ 20.04 GFLOPS | Progress: (8/20) | 7.76 s
[Task 14/25] Current/Best: 7.12/ 20.04 GFLOPS | Progress: (12/20) | 10.51 s
[Task 14/25] Current/Best: 2.83/ 20.04 GFLOPS | Progress: (16/20) | 13.63 s Done.
-
[Task 14/25] Current/Best: 15.05/ 20.04 GFLOPS | Progress: (20/20) | 16.40 s Done.
-
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 6.87/ 22.27 GFLOPS | Progress: (4/20) | 3.46 s
[Task 15/25] Current/Best: 9.51/ 22.27 GFLOPS | Progress: (8/20) | 5.09 s
[Task 15/25] Current/Best: 16.04/ 22.27 GFLOPS | Progress: (12/20) | 7.99 s
[Task 15/25] Current/Best: 22.58/ 22.58 GFLOPS | Progress: (16/20) | 9.61 s
[Task 15/25] Current/Best: 16.56/ 22.58 GFLOPS | Progress: (20/20) | 11.16 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 9.37/ 20.63 GFLOPS | Progress: (4/20) | 3.35 s
[Task 16/25] Current/Best: 15.18/ 20.63 GFLOPS | Progress: (8/20) | 6.58 s
[Task 16/25] Current/Best: 17.73/ 20.63 GFLOPS | Progress: (12/20) | 8.51 s
[Task 16/25] Current/Best: 16.24/ 20.63 GFLOPS | Progress: (16/20) | 10.46 s
[Task 16/25] Current/Best: 17.40/ 20.63 GFLOPS | Progress: (20/20) |
12.64 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 20.61/ 20.61 GFLOPS | Progress: (4/20) | 4.48 s
[Task 17/25] Current/Best: 8.57/ 20.61 GFLOPS | Progress: (8/20) | 8.15 s
[Task 17/25] Current/Best: 9.31/ 21.68 GFLOPS | Progress: (12/20) | 10.74 s
[Task 17/25] Current/Best: 11.62/ 22.63 GFLOPS | Progress: (16/20) | 13.41 s
[Task 17/25] Current/Best: 10.52/ 22.63 GFLOPS | Progress: (20/20) | 16.29 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 12.97/ 17.79 GFLOPS | Progress: (4/20) | 9.14 s
[Task 18/25] Current/Best: 15.96/ 17.81 GFLOPS | Progress: (8/20) | 12.34 s
[Task 18/25] Current/Best: 13.78/ 17.81 GFLOPS | Progress: (12/20) | 15.49 s
[Task 18/25] Current/Best: 11.13/ 17.81 GFLOPS | Progress: (16/20) | 17.74 s
[Task 18/25] Current/Best: 11.27/ 20.92 GFLOPS | Progress: (20/20) | 22.05 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 8.60/ 18.59 GFLOPS | Progress: (4/20) | 6.16 s
[Task 19/25] Current/Best: 10.84/ 18.59 GFLOPS | Progress: (8/20) | 9.15 s
[Task 19/25] Current/Best: 17.37/ 20.62 GFLOPS | Progress: (12/20) | 13.40 s
[Task 19/25] Current/Best: 6.17/ 20.62 GFLOPS | Progress: (16/20) | 15.99 s
[Task 19/25] Current/Best: 10.62/ 20.62 GFLOPS | Progress: (20/20) | 19.29 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 14.50/ 17.46 GFLOPS | Progress: (4/20) | 4.27 s
[Task 20/25] Current/Best: 8.92/ 17.46 GFLOPS | Progress: (8/20) | 6.96 s
[Task 20/25] Current/Best: 18.81/ 18.81 GFLOPS | Progress: (12/20) | 9.85 s
[Task 20/25] Current/Best: 11.78/ 18.81 GFLOPS | Progress: (16/20) | 12.44 s
[Task 20/25] Current/Best: 15.82/ 18.81 GFLOPS | Progress: (20/20) | 15.61 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 8.78/ 12.61 GFLOPS | Progress: (4/20) | 9.28 s
[Task 1/25] Current/Best: 9.93/ 12.61 GFLOPS | Progress: (8/20) | 14.44 s
[Task 1/25] Current/Best: 23.56/ 23.56 GFLOPS | Progress: (12/20) | 16.53 s
[Task 1/25] Current/Best: 15.75/ 23.56 GFLOPS | Progress: (16/20) | 19.19 s
[Task 1/25] Current/Best: 11.81/ 23.72 GFLOPS | Progress: (20/20) | 21.34 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 9.43/ 9.43 GFLOPS | Progress: (4/20) | 3.65 s
[Task 2/25] Current/Best: 20.97/ 20.97 GFLOPS | Progress: (8/20) | 5.21 s
[Task 2/25] Current/Best: 17.13/ 20.97 GFLOPS | Progress: (12/20) | 6.67 s
[Task 2/25] Current/Best: 9.45/ 20.97 GFLOPS | Progress: (16/20) | 8.14 s
[Task 2/25] Current/Best: 14.86/ 21.19 GFLOPS | Progress: (20/20) | 9.52 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 17.83/ 17.83 GFLOPS | Progress: (4/20) | 5.87 s
[Task 3/25] Current/Best: 12.48/ 17.83 GFLOPS | Progress: (8/20) | 8.41 s
[Task 3/25] Current/Best: 6.87/ 17.83 GFLOPS | Progress: (12/20) | 10.87 s
[Task 3/25] Current/Best: 8.33/ 20.35 GFLOPS | Progress: (16/20) | 12.99 s
[Task 3/25] Current/Best: 2.98/ 21.03 GFLOPS | Progress: (20/20) | 15.75 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 13.99/ 16.91 GFLOPS | Progress: (4/20) | 5.78 s
[Task 4/25] Current/Best: 3.32/ 17.62 GFLOPS | Progress: (8/20) | 7.66 s
[Task 4/25] Current/Best: 16.35/ 17.62 GFLOPS | Progress: (12/20) | 9.52 s
[Task 4/25] Current/Best: 10.87/ 17.62 GFLOPS | Progress: (16/20) | 14.34 s
[Task 4/25] Current/Best: 5.25/ 20.61 GFLOPS | Progress: (20/20) | 16.86 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 8.01/ 8.56 GFLOPS | Progress: (4/20) | 4.15 s
[Task 5/25] Current/Best: 9.04/ 13.72 GFLOPS | Progress: (8/20) | 7.08 s
[Task 5/25] Current/Best: 12.63/ 20.74 GFLOPS | Progress: (12/20) | 10.75 s
[Task 5/25] Current/Best: 2.72/ 20.74 GFLOPS | Progress: (16/20) | 12.96 s
[Task 5/25] Current/Best: 5.58/ 20.74 GFLOPS | Progress: (20/20) | 15.52 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 10.10/ 15.73 GFLOPS | Progress: (4/20) | 4.27 s
[Task 6/25] Current/Best: 5.93/ 16.86 GFLOPS | Progress: (8/20) | 7.32 s
[Task 6/25] Current/Best: 13.85/ 17.19 GFLOPS | Progress: (12/20) | 10.02 s
[Task 6/25] Current/Best: 12.90/ 17.19 GFLOPS | Progress: (16/20) | 13.41 s
[Task 6/25] Current/Best: 3.83/ 17.19 GFLOPS | Progress: (20/20) | 17.06 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 11.42/ 11.57 GFLOPS | Progress: (4/20) | 5.91 s
[Task 7/25] Current/Best: 6.20/ 23.01 GFLOPS | Progress: (8/20) | 9.80 s
[Task 7/25] Current/Best: 6.43/ 23.01 GFLOPS | Progress: (12/20) | 12.42 s
[Task 7/25] Current/Best: 7.82/ 23.01 GFLOPS | Progress: (16/20) | 15.07 s
[Task 7/25] Current/Best: 17.15/ 23.01 GFLOPS | Progress: (20/20) | 16.95 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 11.65/ 13.04 GFLOPS | Progress: (4/20) | 8.06 s
[Task 8/25] Current/Best: 13.01/ 13.74 GFLOPS | Progress: (8/20) | 10.50 s
[Task 8/25] Current/Best: 8.18/ 13.74 GFLOPS | Progress: (12/20) | 15.66 s
[Task 8/25] Current/Best: 3.04/ 13.74 GFLOPS | Progress: (16/20) | 20.03 s
[Task 8/25] Current/Best: 14.90/ 14.90 GFLOPS | Progress: (20/20) | 23.85 s Done.
+
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 22.35/ 22.35 GFLOPS | Progress: (4/20) | 7.37 s
[Task 9/25] Current/Best: 6.81/ 22.35 GFLOPS | Progress: (8/20) | 12.37 s
[Task 9/25] Current/Best: 12.97/ 22.35 GFLOPS | Progress: (12/20) | 14.58 s
[Task 9/25] Current/Best: 6.20/ 22.35 GFLOPS | Progress: (16/20) | 16.32 s
[Task 9/25] Current/Best: 5.92/ 22.35 GFLOPS | Progress: (20/20) | 19.82 s Done.
+
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 17.71/ 17.71 GFLOPS | Progress: (4/20) | 3.81 s
[Task 10/25] Current/Best: 3.39/ 17.71 GFLOPS | Progress: (8/20) | 5.90 s
[Task 10/25] Current/Best: 18.96/ 18.96 GFLOPS | Progress: (12/20) | 7.53 s
[Task 10/25] Current/Best: 13.53/ 18.96 GFLOPS | Progress: (16/20) | 9.10 s
[Task 10/25] Current/Best: 18.53/ 18.96 GFLOPS | Progress: (20/20) | 11.11 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 22.63/ 22.63 GFLOPS | Progress: (4/20) | 4.43 s
[Task 11/25] Current/Best: 12.30/ 22.63 GFLOPS | Progress: (8/20) | 6.71 s
[Task 11/25] Current/Best: 9.41/ 22.63 GFLOPS | Progress: (12/20) | 9.49 s
[Task 11/25] Current/Best: 14.93/ 22.63 GFLOPS | Progress: (16/20) | 11.68 s
[Task 11/25] Current/Best: 16.60/ 22.63 GFLOPS | Progress: (20/20) | 14.23 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 12.31/ 14.73 GFLOPS | Progress: (4/20) | 5.88 s
[Task 12/25] Current/Best: 10.32/ 19.72 GFLOPS | Progress: (8/20) | 7.89 s
[Task 12/25] Current/Best: 9.42/ 19.72 GFLOPS | Progress: (12/20) | 10.03 s
[Task 12/25] Current/Best: 7.63/ 19.72 GFLOPS | Progress: (16/20) | 13.76 s
[Task 12/25] Current/Best: 5.14/ 19.72 GFLOPS | Progress: (20/20) | 17.86 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 17.19/ 17.98 GFLOPS | Progress: (4/20) | 6.36 s
[Task 13/25] Current/Best: 18.19/ 18.19 GFLOPS | Progress: (8/20) | 9.59 s
[Task 13/25] Current/Best: 19.38/ 19.38 GFLOPS | Progress: (12/20) | 12.98 s
[Task 13/25] Current/Best: 8.96/ 19.38 GFLOPS | Progress: (16/20) | 15.61 s
[Task 13/25] Current/Best: 6.16/ 19.38 GFLOPS | Progress: (20/20) | 18.09 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 18.11/ 18.11 GFLOPS | Progress: (4/20) | 4.66 s
[Task 14/25] Current/Best: 19.95/ 19.95 GFLOPS | Progress: (8/20) | 7.11 s
[Task 14/25] Current/Best: 9.84/ 19.95 GFLOPS | Progress: (12/20) | 11.96 s
[Task 14/25] Current/Best: 20.69/ 20.69 GFLOPS | Progress: (16/20) | 14.59 s
[Task 14/25] Current/Best: 12.07/ 20.69 GFLOPS | Progress: (20/20) | 19.97 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 9.20/ 22.73 GFLOPS | Progress: (4/20) | 5.79 s
[Task 15/25] Current/Best: 10.81/ 22.73 GFLOPS | Progress: (8/20) | 9.14 s
[Task 15/25] Current/Best: 21.87/ 22.73 GFLOPS | Progress: (12/20) | 12.33 s
[Task 15/25] Current/Best: 3.24/ 22.73 GFLOPS | Progress: (16/20) | 17.40 s
[Task 15/25] Current/Best: 6.91/ 22.73 GFLOPS | Progress: (20/20
) | 19.15 s Done.
+
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 14.21/ 17.51 GFLOPS | Progress: (4/20) | 3.63 s
[Task 16/25] Current/Best: 19.36/ 19.36 GFLOPS | Progress: (8/20) | 5.25 s
[Task 16/25] Current/Best: 17.67/ 19.36 GFLOPS | Progress: (12/20) | 7.02 s
[Task 16/25] Current/Best: 15.17/ 19.36 GFLOPS | Progress: (16/20) | 9.28 s
[Task 16/25] Current/Best: 15.27/ 19.36 GFLOPS | Progress: (20/20) | 10.81 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 19.18/ 19.18 GFLOPS | Progress: (4/20) | 5.48 s
[Task 17/25] Current/Best: 9.31/ 20.94 GFLOPS | Progress: (8/20) | 7.73 s
[Task 17/25] Current/Best: 19.91/ 20.94 GFLOPS | Progress: (12/20) | 9.91 s
[Task 17/25] Current/Best: 15.12/ 20.94 GFLOPS | Progress: (16/20) | 11.93 s
[Task 17/25] Current/Best: 13.17/ 20.94 GFLOPS | Progress: (20/20) | 15.01 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 14.79/ 18.35 GFLOPS | Progress: (4/20) | 3.71 s Done.
+
[Task 18/25] Current/Best: 20.63/ 21.06 GFLOPS | Progress: (8/20) | 6.33 s
[Task 18/25] Current/Best: 19.41/ 21.06 GFLOPS | Progress: (12/20) | 10.98 s
[Task 18/25] Current/Best: 12.87/ 21.06 GFLOPS | Progress: (16/20) | 18.52 s
[Task 18/25] Current/Best: 13.25/ 21.06 GFLOPS | Progress: (20/20) | 22.20 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 4.70/ 17.91 GFLOPS | Progress: (4/20) | 5.89 s
[Task 19/25] Current/Best: 19.92/ 19.92 GFLOPS | Progress: (8/20) | 9.64 s
[Task 19/25] Current/Best: 17.38/ 19.92 GFLOPS | Progress: (12/20) | 13.68 s
[Task 19/25] Current/Best: 1.55/ 21.89 GFLOPS | Progress: (16/20) | 20.38 s
[Task 19/25] Current/Best: 17.52/ 21.99 GFLOPS | Progress: (20/20) | 23.66 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 19.05/ 19.05 GFLOPS | Progress: (4/20) | 4.11 s
[Task 20/25] Current/Best: 12.82/ 19.05 GFLOPS | Progress: (8/20) | 6.68 s
[Task 20/25] Current/Best: 6.13/ 19.05 GFLOPS | Progress: (12/20) | 9.28 s
[Task 20/25] Current/Best: 17.16/ 19.05 GFLOPS | Progress: (16/20) | 12.06 s
[Task 20/25] Current/Best: 15.89/ 19.05 GFLOPS | Progress: (20/20) | 15.24 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 16.42/ 16.42 GFLOPS | Progress: (4/20) | 4.04 s
[Task 21/25] Current/Best: 11.27/ 20.90 GFLOPS | Progress: (8/20) | 8.11 s
[Task 21/25] Current/Best: 5.28/ 20.90 GFLOPS | Progress: (12/20) | 11.55 s
[Task 21/25] Current/Best: 15.32/ 20.90 GFLOPS | Progress: (16/20) | 13.97 s
[Task 21/25] Current/Best: 15.98/ 20.90 GFLOPS | Progress: (20/20)
| 15.59 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
Done.
-
[Task 21/25] Current/Best: 20.50/ 20.50 GFLOPS | Progress: (4/20) | 3.46 s
[Task 21/25] Current/Best: 9.60/ 20.50 GFLOPS | Progress: (8/20) | 5.61 s
[Task 21/25] Current/Best: 4.48/ 20.50 GFLOPS | Progress: (12/20) | 8.05 s
[Task 21/25] Current/Best: 10.69/ 20.90 GFLOPS | Progress: (16/20) | 11.18 s
[Task 21/25] Current/Best: 21.33/ 21.33 GFLOPS | Progress: (20/20) | 13.13 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 9.53/ 15.01 GFLOPS | Progress: (4/20) | 4.30 s
[Task 22/25] Current/Best: 6.23/ 16.00 GFLOPS | Progress: (8/20) | 6.52 s
[Task 22/25] Current/Best: 15.06/ 16.00 GFLOPS | Progress: (12/20) | 10.04 s
[Task 22/25] Current/Best: 10.86/ 19.94 GFLOPS | Progress: (16/20) | 12.28 s
[Task 22/25] Current/Best: 9.58/ 19.94 GFLOPS | Progress: (20/20) | 15.10 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 10.84/ 19.13 GFLOPS | Progress: (4/20) | 4.41 s
[Task 23/25] Current/Best: 10.33/ 22.16 GFLOPS | Progress: (8/20) | 7.38 s
[Task 23/25] Current/Best: 12.22/ 23.73 GFLOPS | Progress: (12/20) | 9.57 s
[Task 23/25] Current/Best: 9.66/ 23.73 GFLOPS | Progress: (16/20) | 13.19 s
[Task 23/25] Current/Best: 3.09/ 23.73 GFLOPS | Progress: (20/20) | 16.15 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 2.85/ 5.45 GFLOPS | Progress: (4/20) | 11.16 s
[Task 24/25] Current/Best: 4.25/ 5.45 GFLOPS | Progress: (8/20) | 14.90 s
[Task 24/25] Current/Best: 3.13/ 5.45 GFLOPS | Progress: (12/20) | 25.87 s
[Task 24/25] Current/Best: 1.08/ 10.56 GFLOPS | Progress: (16/20) | 36.81 s
[Task 24/25] Current/Best: 6.99/ 10.56 GFLOPS | Progress: (20/20) | 47.83 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 25/25] Current/Best: 5.35/ 6.55 GFLOPS | Progress: (4/20) | 12.73 s
[Task 25/25] Current/Best: 9.41/ 9.41 GFLOPS | Progress: (8/20) | 15.60 s
[Task 25/25] Current/Best: 3.76/ 9.41 GFLOPS | Progress: (12/20) | 26.55 s
[Task 25/25] Current/Best: 3.59/ 9.41 GFLOPS | Progress: (16/20) | 38.32 s
[Task 25/25] Current/Best: 1.55/ 9.41 GFLOPS | Progress: (20/20) | 49.82 s
+
[Task 22/25] Current/Best: 2.69/ 13.91 GFLOPS | Progress: (4/20) | 4.76 s
[Task 22/25] Current/Best: 18.66/ 18.66 GFLOPS | Progress: (8/20) | 6.69 s
[Task 22/25] Current/Best: 12.95/ 18.66 GFLOPS | Progress: (12/20) | 9.24 s
[Task 22/25] Current/Best: 10.33/ 18.66 GFLOPS | Progress: (16/20) | 11.76 s
[Task 22/25] Current/Best: 19.38/ 19.38 GFLOPS | Progress: (20/20) | 14.48 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 17.50/ 17.50 GFLOPS | Progress: (4/20) | 9.48 s
[Task 23/25] Current/Best: 16.00/ 19.52 GFLOPS | Progress: (8/20) | 15.30 s
[Task 23/25] Current/Best: 4.65/ 19.52 GFLOPS | Progress: (12/20) | 18.20 s
[Task 23/25] Current/Best: 10.56/ 19.52 GFLOPS | Progress: (16/20) | 20.69 s
[Task 23/25] Current/Best: 18.40/ 19.52 GFLOPS | Progress: (20/20) | 24.15 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 8.23/ 8.23 GFLOPS | Progress: (4/20) | 5.57 s
[Task 24/25] Current/Best: 7.09/ 8.23 GFLOPS | Progress: (8/20) | 16.49 s
[Task 24/25] Current/Best: 6.76/ 8.23 GFLOPS | Progress: (12/20) | 21.27 s
[Task 24/25] Current/Best: 3.53/ 8.23 GFLOPS | Progress: (16/20) | 32.22 s
[Task 24/25] Current/Best: 1.73/ 9.98 GFLOPS | Progress: (20/20) | 34.76 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 3.49/ 9.25 GFLOPS | Progress: (4/20) | 3.20 s
[Task 25/25] Current/Best: 2.90/ 9.33 GFLOPS | Progress: (8/20) | 9.27 s
[Task 25/25] Current/Best: 7.49/ 9.33 GFLOPS | Progress: (12/20) | 11.31 s
[Task 25/25] Current/Best: 2.99/ 9.33 GFLOPS | Progress: (16/20) | 12.70 s
[Task 25/25] Current/Best: 7.38/ 9.33 GFLOPS | Progress: (20/2
0) | 23.65 s
@@ -722,8 +722,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 423.99232671999926, 'median': 424.4845817998794, 'std': 1.2220797649777073}
- unoptimized: {'mean': 520.7501513800526, 'median': 520.3653747001226, 'std': 2.5685746603934763}
+ optimized: {'mean': 428.98721735999516, 'median': 428.70022719998815, 'std': 3.5759263153718526}
+ unoptimized: {'mean': 523.5642570399989, 'median': 524.7381491499993, 'std': 3.304938946858366}
@@ -746,7 +746,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 11 minutes 31.989 seconds)
+ **Total running time of the script:** ( 11 minutes 23.703 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 da52c3f959..24009f1990 100644
--- a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
+++ b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
@@ -274,7 +274,7 @@ device and returns the measured cost. Network overhead is excluded.
.. code-block:: none
- 2.266e-07 secs/op
+ 1.319e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index 39b9731102..631556e76a 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -263,7 +263,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x1a5670c0)), stage(b, placeholder(b, 0x1ba83c70)), 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, 0x4d869c0)), stage(b, placeholder(b, 0xc110e40)), stage(T_add, compute(T_add, body=[a[ax0, ax1, ax2] + b[ax1, ax2]], axis=[T.iter_var(ax0, T.Range(0, 100), "DataPar", ""), T.iter_var(ax1, T.Range(0, 10), "DataPar", ""), T.iter_var(ax2, T.Range(0, 10), "DataPar", "")], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[a[ax0, ax1, ax2] * b[ax1, ax2]], axis=[T.iter_var(ax0, T.Range(0, 100), "DataPar", ""), T.iter_var(ax1, T.R [...]
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index dbd9032ba0..26cccea12c 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,32 +5,32 @@
Computation times
=================
-**14:48.385** total execution time for **tutorial** files:
+**15:16.681** total execution time for **tutorial** files:
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 11:31.989 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 11:23.703 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:14.772 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:40.995 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 00:59.094 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:00.042 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:33.351 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:35.859 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:27.600 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:33.787 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.814 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.279 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:00.602 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.832 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.164 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.184 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_uma.py` (``uma.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.000 | 0.0 MB |
-+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_tvmc_python.py` (``tvmc_python.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
+| :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``) | 00:00.000 | 0.0 MB |
++------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_install.py` (``install.py``) | 00:00.000 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
index 70be0065b3..5c074c6985 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -286,7 +286,7 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
Numpy running time: 0.000008
- naive: 0.000009
+ naive: 0.000007
@@ -490,10 +490,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 8.092030002444517e-06 1.0
- naive 9.287e-06 1.1476724625581585
- parallel 7.241000000000001e-06 0.8948310866139363
- vector 2.4590100000000004e-05 3.0388048478035294
+ numpy 7.650969998849177e-06 1.0
+ naive 6.7019999999999995e-06 0.8759673611330436
+ parallel 6.951599999999999e-06 0.9085906755673628
+ vector 2.4604e-05 3.215801395600926
@@ -914,7 +914,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.018492
+ Numpy running time: 0.019110
@@ -972,7 +972,7 @@ optimizations.
.. code-block:: none
- none: 3.265450
+ none: 3.276493
@@ -1074,7 +1074,7 @@ schedule.
.. code-block:: none
- blocking: 0.300937
+ blocking: 0.327767
@@ -1169,7 +1169,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.339822
+ vectorization: 0.355377
@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, [1024, 1024], []),
@@ -1242,7 +1242,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.112812
+ loop permutation: 0.126815
@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, [1024, 1024], []),
@@ -1340,7 +1340,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.107125
+ array packing: 0.111627
@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, [1024, 1024], []),
@@ -1432,7 +1432,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.113946
+ block caching: 0.111632
@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, [1024, 1024], []),
@@ -1517,7 +1517,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.149286
+ parallelization: 0.146850
@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, [1024, 1024], []),
@@ -1597,13 +1597,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.265449653 1.0
- blocking 0.30093717870000003 0.09215796006027106
- vectorization 0.3398221805 0.10406596843035142
- loop permutation 0.11281171059999999 0.034547067812348224
- array packing 0.10712526550000001 0.03280567054573418
- block caching 0.1139461139 0.03489446355276573
- parallelization 0.1492855438 0.04571668825542906
+ none 3.2764933225 1.0
+ blocking 0.3277672024 0.10003597448198372
+ vectorization 0.3553773964 0.10846272567063961
+ loop permutation 0.1268146886 0.03870439403283722
+ array packing 0.1116269298 0.03406902404880454
+ block caching 0.1116318513 0.03407052611199088
+ parallelization 0.1468502481 0.044819333856585326
@@ -1643,6 +1643,11 @@ 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 0.042 seconds)
+
+
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
.. only:: html
diff --git a/docs/commit_hash b/docs/commit_hash
index 9b8c2aeee3..d84ab2184e 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-60c723ec267ee5095a35add5f8259e650b8ddd7b
+c452e6966c33047512155f42f63aef4e0586d129
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index 14abd2b553..51391faafe 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -585,7 +585,7 @@ class:['truck 0.9266'] left:471 top:83 right:689 bottom:169
class:['bicycle 0.9984'] left:111 top:113 right:577 bottom:447
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 8.187 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 15.348 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-darknet-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/7716f96385bd5abb6e822041e285be54/from_darknet.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_darknet.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/from_keras.html b/docs/how_to/compile_models/from_keras.html
index 5e619c038e..09bd4103ee 100644
--- a/docs/how_to/compile_models/from_keras.html
+++ b/docs/how_to/compile_models/from_keras.html
@@ -506,7 +506,7 @@ Tensorflow is also required since it’s used as the default backend of keras.</
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Relay top-1 id: 285, class name: Egyptian cat
1/1 [==============================] - ETA: 0s
-1/1 [==============================] - 1s 930ms/step
+1/1 [==============================] - 1s 947ms/step
Keras top-1 id: 285, class name: Egyptian cat
</pre></div>
</div>
diff --git a/docs/how_to/compile_models/from_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index 348b9ecd05..47fbdca464 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -439,7 +439,7 @@
<span class="nb">print</span><span class="p">(</span><span class="s2">"x"</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#tuple" title="builtins.tuple" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">x</span><span class="o">.</span><span class="n">shape</span></a><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip3df8cdd9-4f05-4c71-a420-c4b44ef8ddb3 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipc7c73fb1-e93f-4182-85ca-fb0516cad9c0 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 84fdbf22a9..97ca98268c 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -449,13 +449,12 @@ Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdo
<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]
- 15%|#5 | 6.33M/41.5M [00:00<00:00, 43.2MB/s]
- 25%|##5 | 10.5M/41.5M [00:00<00:01, 27.4MB/s]
- 41%|#### | 17.0M/41.5M [00:00<00:00, 39.9MB/s]
- 58%|#####7 | 24.0M/41.5M [00:00<00:00, 37.2MB/s]
- 78%|#######8 | 32.5M/41.5M [00:00<00:00, 49.7MB/s]
- 92%|#########2| 38.3M/41.5M [00:00<00:00, 48.1MB/s]
-100%|##########| 41.5M/41.5M [00:01<00:00, 43.3MB/s]
+ 19%|#9 | 7.99M/41.5M [00:00<00:00, 45.9MB/s]
+ 39%|###8 | 16.0M/41.5M [00:00<00:00, 61.9MB/s]
+ 58%|#####7 | 24.0M/41.5M [00:00<00:00, 68.0MB/s]
+ 77%|#######7 | 32.0M/41.5M [00:00<00:00, 67.7MB/s]
+ 96%|#########6| 40.0M/41.5M [00:00<00:00, 71.2MB/s]
+100%|##########| 41.5M/41.5M [00:00<00:00, 68.8MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index 4606708232..1ec15cf3db 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -432,12 +432,10 @@ be unstable.</p>
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]
- 18%|#7 | 7.99M/44.7M [00:00<00:00, 54.6MB/s]
- 36%|###5 | 16.0M/44.7M [00:00<00:00, 57.8MB/s]
- 54%|#####3 | 24.0M/44.7M [00:00<00:00, 59.2MB/s]
- 72%|#######1 | 32.0M/44.7M [00:00<00:00, 57.0MB/s]
- 88%|########8 | 39.3M/44.7M [00:00<00:00, 62.5MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 60.6MB/s]
+ 30%|### | 13.6M/44.7M [00:00<00:00, 141MB/s]
+ 61%|###### | 27.0M/44.7M [00:00<00:00, 93.6MB/s]
+ 90%|########9 | 40.0M/44.7M [00:00<00:00, 106MB/s]
+100%|##########| 44.7M/44.7M [00:00<00:00, 106MB/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 54cb2ba082..e108d67faf 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -649,7 +649,7 @@ 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 10.219 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 20.833 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-tensorflow-py">
<div class="sphx-glr-download sphx-glr-download-python 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 3c995d886a..d56b9696af 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>05:37.346</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>06:20.885</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 81%" />
@@ -349,43 +349,43 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>01:10.219</p></td>
+<td><p>01:20.833</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>01:08.187</p></td>
+<td><p>01:15.348</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:45.442</p></td>
+<td><p>00:53.068</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:30.946</p></td>
+<td><p>00:35.786</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:28.392</p></td>
+<td><p>00:30.700</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:25.999</p></td>
+<td><p>00:30.611</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:25.325</p></td>
+<td><p>00:26.729</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:23.244</p></td>
+<td><p>00:24.782</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:17.145</p></td>
+<td><p>00:20.421</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:02.446</p></td>
+<td><p>00:02.607</p></td>
<td><p>0.0 MB</p></td>
</tr>
</tbody>
diff --git a/docs/how_to/deploy_models/deploy_model_on_adreno.html b/docs/how_to/deploy_models/deploy_model_on_adreno.html
index e8c9cc6701..2b2c702ed3 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno.html
@@ -920,7 +920,7 @@ Top5 predictions:
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2751.7866 2751.6081 2754.8158 2749.1211 1.9819
+ 2755.3963 2755.1467 2757.3894 2753.0723 1.4283
</pre></div>
</div>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-model-on-adreno-py">
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 4665177a34..9a4fbbe935 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -662,7 +662,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 17.1610 16.8567 21.1197 16.0662 1.4243
+ 16.1355 16.1322 16.2570 15.9628 0.0838
</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 6660b632bb..0404dfde13 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -454,27 +454,24 @@ be unstable.</p>
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]
- 8%|8 | 14.0M/170M [00:00<00:01, 147MB/s]
- 16%|#6 | 28.0M/170M [00:00<00:01, 76.8MB/s]
- 22%|##1 | 37.0M/170M [00:00<00:02, 63.3MB/s]
- 26%|##5 | 43.9M/170M [00:00<00:02, 64.8MB/s]
- 30%|##9 | 50.8M/170M [00:00<00:02, 61.0MB/s]
- 34%|###3 | 56.9M/170M [00:00<00:02, 58.2MB/s]
- 38%|###7 | 64.0M/170M [00:01<00:01, 59.4MB/s]
- 42%|####2 | 72.1M/170M [00:01<00:01, 65.9MB/s]
- 47%|####7 | 80.0M/170M [00:01<00:01, 59.9MB/s]
- 52%|#####1 | 88.0M/170M [00:01<00:01, 62.0MB/s]
- 57%|#####6 | 96.0M/170M [00:01<00:01, 62.6MB/s]
- 61%|######1 | 104M/170M [00:01<00:01, 60.8MB/s]
- 66%|######5 | 112M/170M [00:01<00:00, 64.4MB/s]
- 73%|#######3 | 125M/170M [00:01<00:00, 81.7MB/s]
- 79%|#######8 | 134M/170M [00:02<00:00, 84.4MB/s]
- 84%|########3 | 142M/170M [00:02<00:00, 79.7MB/s]
- 88%|########8 | 150M/170M [00:02<00:00, 73.9MB/s]
- 92%|#########2| 157M/170M [00:02<00:00, 68.8MB/s]
- 96%|#########6| 164M/170M [00:02<00:00, 60.4MB/s]
-100%|#########9| 170M/170M [00:02<00:00, 57.8MB/s]
-100%|##########| 170M/170M [00:02<00:00, 66.2MB/s]
+ 5%|4 | 7.99M/170M [00:00<00:02, 66.6MB/s]
+ 9%|9 | 16.0M/170M [00:00<00:02, 69.8MB/s]
+ 14%|#4 | 24.0M/170M [00:00<00:02, 67.7MB/s]
+ 20%|## | 34.1M/170M [00:00<00:01, 81.0MB/s]
+ 25%|##4 | 42.1M/170M [00:00<00:01, 78.8MB/s]
+ 29%|##9 | 49.7M/170M [00:00<00:01, 73.7MB/s]
+ 34%|###4 | 57.9M/170M [00:00<00:01, 77.3MB/s]
+ 40%|#### | 68.4M/170M [00:00<00:01, 87.0MB/s]
+ 45%|####5 | 76.9M/170M [00:01<00:01, 86.9MB/s]
+ 52%|#####1 | 87.6M/170M [00:01<00:00, 94.6MB/s]
+ 57%|#####6 | 96.8M/170M [00:01<00:01, 75.6MB/s]
+ 64%|######4 | 109M/170M [00:01<00:00, 89.6MB/s]
+ 71%|####### | 120M/170M [00:01<00:00, 92.7MB/s]
+ 79%|#######8 | 134M/170M [00:01<00:00, 106MB/s]
+ 85%|########4 | 144M/170M [00:01<00:00, 97.1MB/s]
+ 91%|######### | 154M/170M [00:01<00:00, 96.9MB/s]
+ 97%|#########7| 165M/170M [00:01<00:00, 102MB/s]
+100%|##########| 170M/170M [00:02<00:00, 88.9MB/s]
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: 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)
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: 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=& [...]
@@ -572,7 +569,7 @@ 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> ( 3 minutes 10.544 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 36.065 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-object-detection-pytorch-py">
<div class="sphx-glr-download sphx-glr-download-python 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 b38039dd55..43d510a572 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -495,8 +495,8 @@ training. Other models require a full post training calibration.</p>
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]
- 59%|#####8 | 7.99M/13.6M [00:00<00:00, 75.0MB/s]
-100%|##########| 13.6M/13.6M [00:00<00:00, 84.3MB/s]
+ 59%|#####8 | 7.99M/13.6M [00:00<00:00, 64.7MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 91.5MB/s]
</pre></div>
</div>
</div>
@@ -587,7 +587,7 @@ output values are identical out of 1000 outputs from mobilenet v2.</p>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 90.4797 90.3710 94.1902 90.1895 0.4371
+ 90.3827 90.3425 90.7219 90.0761 0.1844
</pre></div>
</div>
<div class="admonition note">
@@ -626,7 +626,7 @@ 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 5.071 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 14.229 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-py">
<div class="sphx-glr-download sphx-glr-download-python 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 ef833ae3f0..9db6299679 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -580,7 +580,7 @@ TFLite Top-5 labels: [387 102 386 341 349]
</div>
<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)
- 119.4774 119.3662 124.0014 117.3927 1.1786
+ 121.0030 120.9543 125.1200 119.8180 0.7057
</pre></div>
</div>
<div class="admonition note">
@@ -608,7 +608,7 @@ 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> ( 2 minutes 29.124 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 33.566 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-tflite-py">
<div class="sphx-glr-download sphx-glr-download-python 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 9a235ecfa6..8e65930796 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -521,7 +521,7 @@ 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 50.139 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 30.231 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-quantized-py">
<div class="sphx-glr-download sphx-glr-download-python 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 2efe7b1435..7dfc785dae 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -463,22 +463,25 @@ 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]
- 4%|4 | 5877/132723 [00:00<00:02, 58765.71KB/s]
- 11%|#1 | 14645/132723 [00:00<00:01, 75770.04KB/s]
- 18%|#7 | 23350/132723 [00:00<00:01, 80918.32KB/s]
- 24%|##4 | 32022/132723 [00:00<00:01, 83200.24KB/s]
- 31%|### | 40841/132723 [00:00<00:01, 84990.39KB/s]
- 37%|###7 | 49629/132723 [00:00<00:00, 85970.32KB/s]
- 44%|####4 | 58440/132723 [00:00<00:00, 86665.60KB/s]
- 51%|##### | 67188/132723 [00:00<00:00, 86923.48KB/s]
- 57%|#####7 | 75987/132723 [00:00<00:00, 87255.07KB/s]
- 64%|######3 | 84900/132723 [00:01<00:00, 87830.65KB/s]
- 71%|####### | 93784/132723 [00:01<00:00, 88138.60KB/s]
- 77%|#######7 | 102598/132723 [00:01<00:00, 88111.64KB/s]
- 84%|########3 | 111483/132723 [00:01<00:00, 88333.37KB/s]
- 91%|######### | 120324/132723 [00:01<00:00, 88354.31KB/s]
- 97%|#########7| 129196/132723 [00:01<00:00, 88456.73KB/s]
-100%|##########| 132723/132723 [00:01<00:00, 86088.15KB/s]
+ 3%|2 | 3802/132723 [00:00<00:03, 38007.17KB/s]
+ 8%|7 | 10553/132723 [00:00<00:02, 55357.30KB/s]
+ 13%|#3 | 17833/132723 [00:00<00:01, 63314.99KB/s]
+ 19%|#9 | 25223/132723 [00:00<00:01, 67491.97KB/s]
+ 25%|##4 | 32549/132723 [00:00<00:01, 69568.22KB/s]
+ 30%|### | 39915/132723 [00:00<00:01, 70956.56KB/s]
+ 36%|###5 | 47300/132723 [00:00<00:01, 71900.70KB/s]
+ 41%|####1 | 54627/132723 [00:00<00:01, 72334.79KB/s]
+ 47%|####6 | 61862/132723 [00:00<00:00, 72334.56KB/s]
+ 52%|#####2 | 69197/132723 [00:01<00:00, 72640.60KB/s]
+ 58%|#####7 | 76578/132723 [00:01<00:00, 72995.81KB/s]
+ 63%|######3 | 83988/132723 [00:01<00:00, 73330.31KB/s]
+ 69%|######8 | 91322/132723 [00:01<00:00, 73231.57KB/s]
+ 74%|#######4 | 98646/132723 [00:01<00:00, 73201.85KB/s]
+ 80%|#######9 | 105998/132723 [00:01<00:00, 73293.99KB/s]
+ 85%|########5 | 113345/132723 [00:01<00:00, 73345.21KB/s]
+ 91%|######### | 120680/132723 [00:01<00:00, 73279.05KB/s]
+ 96%|#########6| 128008/132723 [00:01<00:00, 73209.82KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 71088.57KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -517,7 +520,7 @@ Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from h
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 2.230 seconds)</p>
+<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 34.278 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-ssd-gluoncv-py">
<div class="sphx-glr-download sphx-glr-download-python 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 6f0adf6236..67cf4c0d0a 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>13:55.486</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>15:01.388</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 86%" />
@@ -349,39 +349,39 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>03:10.544</p></td>
+<td><p>03:36.065</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>03:02.230</p></td>
+<td><p>03:34.278</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>02:29.124</p></td>
+<td><p>02:33.566</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>01:50.139</p></td>
+<td><p>01:30.231</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>01:05.071</p></td>
+<td><p>01:14.229</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_adreno.html#sphx-glr-how-to-deploy-models-deploy-model-on-adreno-py"><span class="std std-ref">Deploy the Pretrained Model on Adreno</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_adreno.py</span></code>)</p></td>
-<td><p>00:53.556</p></td>
+<td><p>00:56.417</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:34.648</p></td>
+<td><p>00:40.938</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_nano.html#sphx-glr-how-to-deploy-models-deploy-model-on-nano-py"><span class="std std-ref">Deploy the Pretrained Model on Jetson Nano</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_nano.py</span></code>)</p></td>
-<td><p>00:25.234</p></td>
+<td><p>00:28.076</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:24.933</p></td>
+<td><p>00:27.582</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
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 91ec124c42..7586066553 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -619,7 +619,7 @@ In this alpha state of the Bring Your Own Datatypes framework, we have not imple
<span class="n">module</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#dict" title="builtins.dict" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">params</span></a> <span class="o">=</span> <span class="n">get_mobilenet</span><span class="p">()</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip01d1a502-6542-48df-a991-b0312e20263c 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.zipaca19498-ab89-4cdf-8bb5-543eb74514c2 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 00d62ea093..8f08f106e9 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>00:46.908</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:53.834</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:43.561</p></td>
+<td><p>00:50.047</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:02.337</p></td>
+<td><p>00:02.694</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:01.004</p></td>
+<td><p>00:01.086</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
diff --git a/docs/how_to/extend_tvm/use_pass_instrument.html b/docs/how_to/extend_tvm/use_pass_instrument.html
index 49b67722d5..f6887efbf2 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -526,10 +526,10 @@ profile the execution time of each passes.</p>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 7368us [7368us] (47.01%; 47.01%)
-FoldScaleAxis: 8304us [6us] (52.99%; 52.99%)
- FoldConstant: 8299us [1679us] (52.95%; 99.93%)
- InferType: 6620us [6620us] (42.24%; 79.77%)
+InferType: 18502us [18502us] (48.95%; 48.95%)
+FoldScaleAxis: 19294us [9us] (51.05%; 51.05%)
+ FoldConstant: 19285us [1711us] (51.02%; 99.95%)
+ InferType: 17574us [17574us] (46.50%; 91.13%)
</pre></div>
</div>
</div>
@@ -551,10 +551,10 @@ Refer to following sections and <a class="reference internal" href="../../refere
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 6620us [6620us] (44.90%; 44.90%)
-FoldScaleAxis: 8124us [4us] (55.10%; 55.10%)
- FoldConstant: 8120us [1655us] (55.07%; 99.95%)
- InferType: 6465us [6465us] (43.84%; 79.62%)
+InferType: 17911us [17911us] (47.93%; 47.93%)
+FoldScaleAxis: 19456us [8us] (52.07%; 52.07%)
+ FoldConstant: 19448us [1782us] (52.05%; 99.96%)
+ InferType: 17667us [17667us] (47.28%; 90.84%)
</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 113f9d0a23..e750511bf8 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -575,7 +575,7 @@ latency of convolution.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Convolution: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">*</span> <span cl [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.238945 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 42.309631 ms
</pre></div>
</div>
<div class="sphx-glr-footer 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 982b11dec8..5f796e33ff 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -915,7 +915,7 @@ be able to run on our build server</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"conv2d with tensor core: </span><span class="si">%f</span><span class="s2"> ms"</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">* [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 8.092464 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 13.360538 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 bee3fa840f..14b6b70ff1 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -472,8 +472,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
<span class="nb">print</span><span class="p">(</span><span class="s2">"Baseline: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018495
-Baseline: 3.193068
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.019325
+Baseline: 3.239618
</pre></div>
</div>
<p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -532,7 +532,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt1: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.300526
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.331288
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -598,7 +598,7 @@ vastly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt2: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.332444
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.344937
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -658,7 +658,7 @@ the access pattern for A matrix is more cache friendly.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt3: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.113062
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.128089
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -740,7 +740,7 @@ flattening.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt4: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.110256
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.110026
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -825,7 +825,7 @@ write to C when all the block results are ready.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt5: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.114555
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.112031
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -914,7 +914,7 @@ write to C when all the block results are ready.</p>
<span class="nb">print</span><span class="p">(</span><span class="s2">"Opt6: </span><span class="si">%f</span><span class="s2">"</span> <span class="o">%</span> <span class="n">opt6_time</span><span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.152187
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.146417
</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 a0f9ca2155..ccf1cc7887 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -340,7 +340,7 @@
<div class="section" id="computation-times">
<span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:34.650</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:35.210</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:31.892</p></td>
+<td><p>00:32.575</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:01.578</p></td>
+<td><p>00:01.552</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:01.180</p></td>
+<td><p>00:01.083</p></td>
<td><p>0.0 MB</p></td>
</tr>
</tbody>
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 bd8a119c28..c404f34ba4 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>08:59.047</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>09:27.857</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 85%" />
@@ -349,27 +349,27 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>05:35.421</p></td>
+<td><p>05:44.391</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>01:30.882</p></td>
+<td><p>01:40.523</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>01:01.117</p></td>
+<td><p>01:06.185</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:28.896</p></td>
+<td><p>00:29.281</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:11.819</p></td>
+<td><p>00:14.284</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:10.913</p></td>
+<td><p>00:13.192</p></td>
<td><p>0.0 MB</p></td>
</tr>
</tbody>
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 0c9385c143..69ed3d021f 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
@@ -504,686 +504,51 @@ cooperative fetching, unrolling and operator fusion.</p>
bias: Buffer(bias_2: Pointer(float32), float32, [1, 512, 1, 1], []),
compute: Buffer(compute_2: Pointer(float32), float32, [1, 512, 7, 7], [])}
buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
- attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 28;
- allocate(conv2d_nchw: Pointer(local float32), float32, [7]), storage_scope = local;
- allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
- allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
- attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128 {
- conv2d_nchw_1: Buffer(conv2d_nchw, float32, [7], [], scope="local", align=16)[0] = 0f32
+ attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 32;
+ allocate(conv2d_nchw: Pointer(local float32), float32, [2]), storage_scope = local;
+ allocate(pad_temp.shared: Pointer(shared float32), float32, [252]), storage_scope = shared;
+ allocate(kernel.shared: Pointer(shared float32), float32, [192]), storage_scope = shared;
+ attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392 {
+ conv2d_nchw_1: Buffer(conv2d_nchw, float32, [1], [], scope="local", align=4)[0] = 0f32
conv2d_nchw_1[1] = 0f32
- conv2d_nchw_1[2] = 0f32
- conv2d_nchw_1[3] = 0f32
- conv2d_nchw_1[4] = 0f32
- conv2d_nchw_1[5] = 0f32
- conv2d_nchw_1[6] = 0f32
- for (rc.outer.outer: int32, 0, 64) {
- let cse_var_1: int32 = (rc.outer.outer*72)
- {
- attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else((((1 <= floormod(blockIdx.x, 7)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) - 8)], 0f32, dtype=float32)
+ for (rc.outer.outer: int32, 0, 128) {
+ for (ry.outer.outer: int32, 0, 3) {
+ attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392 {
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1: Buffer(pad_temp.shared, float32, [252], [], scope="shared")[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1*4), 63), 9) + ry.outer.outer)) && ((floordiv(floormod((threadIdx.x_1*4), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod((threadIdx.x_1*4), 9))) && (floormod((threadIdx.x_1*4), 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((rc.outer.outer*196) + (floordiv((t [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 1), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 1), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 1), 9))) && (floormod(((threadIdx.x_1*4) + 1), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 1), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1* [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 2), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 2), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 2), 9))) && (floormod(((threadIdx.x_1*4) + 2), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 2), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1* [...]
+ }
+ if @tir.likely((threadIdx.x_1 < 63), dtype=bool) {
+ pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 <= (floordiv(floormod(((threadIdx.x_1*4) + 3), 63), 9) + ry.outer.outer)) && ((floordiv(floormod(((threadIdx.x_1*4) + 3), 63), 9) + ry.outer.outer) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 3), 9))) && (floormod(((threadIdx.x_1*4) + 3), 9) < 8)), data_3[(((((rc.outer.outer*196) + (floordiv(((threadIdx.x_1*4) + 3), 9)*7)) + (ry.outer.outer*7)) + floormod(((threadIdx.x_1* [...]
+ }
}
- attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73728)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147456)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221184)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294912)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368640)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442368)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516096)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3))]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3))]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else(((1 <= floormod(threadIdx.x_1, 9)) && (floormod(threadIdx.x_1, 9) < 8)), data_3[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) - 1)], 0f32, dtype=float32)
+ attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 392;
+ if @tir.likely((threadIdx.x_2 < 192), dtype=bool) {
+ kernel.shared_1: Buffer(kernel.shared, float32, [192], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((((blockIdx.x*73728) + (floordiv(threadIdx.x_2, 12)*4608)) + (rc.outer.outer*36)) + (floordiv(floormod(threadIdx.x_2, 12), 3)*9)) + (ry.outer.outer*3)) + floormod(threadIdx.x_2, 3))]
}
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[threadIdx.x_2] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73731)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147459)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221187)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294915)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368643)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442371)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516099)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 3)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 3)]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- if @tir.likely((threadIdx.x_1 < 72), dtype=bool) {
- pad_temp.shared_1[threadIdx.x_1] = @tir.if_then_else((((floormod(blockIdx.x, 7) < 6) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3[(((((rc.outer.outer*392) + (floordiv(threadIdx.x_1, 9)*49)) + (floormod(blockIdx.x, 7)*7)) + floormod(threadIdx.x_1, 9)) + 6)], 0f32, dtype=float32)
+ for (rc.outer.inner: int32, 0, 2) {
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[(((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7))]*kernel.shared_1[((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6))]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[(((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7))]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 96)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 3)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 63)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 99)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 1)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 1)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 97)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 64)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 4)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 64)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 100)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 2)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 2)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 98)]))
+ conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 65)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 5)]))
+ conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[((((rc.outer.inner*126) + (floordiv(floormod(threadIdx.x, 49), 7)*9)) + floormod(threadIdx.x, 7)) + 65)]*kernel.shared_1[(((floordiv(threadIdx.x, 49)*12) + (rc.outer.inner*6)) + 101)]))
}
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[threadIdx.x_2] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 73734)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 147462)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 221190)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 294918)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 368646)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 442374)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_1) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + floormod(threadIdx.x_2, 3)) + 516102)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + floormod((threadIdx.x_2 + 2), 3)) + 6)]
- attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 128;
- kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + floormod((threadIdx.x_2 + 1), 3)) + 6)]
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*24)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 3)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 1)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 4)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*24) + 2)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*24) + 5)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 6)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 9)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 7)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 10)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*24) + 8)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*24) + 11)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 12)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 15)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 13)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 16)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*24) + 14)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*24) + 17)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 18)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 21)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 19)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 22)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*24) + 20)]))
- conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*24) + 23)]))
- conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*24) + 23)]))
}
}
- for (i3.inner: int32, 0, 7) {
- compute_3: Buffer(compute_2, float32, [25088], [])[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[i3.inner] + bias_3: Buffer(bias_2, float32, [512], [])[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
- }
+ compute_3: Buffer(compute_2, float32, [25088], [])[((blockIdx.x*784) + threadIdx.x)] = max((conv2d_nchw_1[0] + bias_3: Buffer(bias_2, float32, [512], [])[((blockIdx.x*16) + floordiv(threadIdx.x, 49))]), 0f32)
+ compute_3[(((blockIdx.x*784) + threadIdx.x) + 392)] = max((conv2d_nchw_1[1] + bias_3[(((blockIdx.x*16) + floordiv(threadIdx.x, 49)) + 8)]), 0f32)
}
}
</pre></div>
@@ -1219,7 +584,7 @@ cooperative fetching, unrolling and operator fusion.</p>
<span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.322 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.404 ms
</pre></div>
</div>
</div>
@@ -1250,18 +615,18 @@ conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_
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=1)
-conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=128)
-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_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=8)
+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_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=7)
+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=2)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, 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=1)
@@ -1271,13 +636,13 @@ 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=1)
-compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=128)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=8)
+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_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=7)
-compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
+compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
+compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=7)
compute_i3_o_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)
@@ -1297,14 +662,14 @@ s[compute].bind(compute_i0_o_i_i1_o_i_fused_i2_o_i_fused_i3_o_i_fused, te.thread
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)
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=128)
+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=392)
s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
+pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=128)
+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=392)
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", 1024)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 16)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -1322,613 +687,49 @@ CUDA source code:
#define int64_t long long
#define uint64_t unsigned long long
#endif
-extern "C" __global__ void __launch_bounds__(128) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[7];
- __shared__ float pad_temp_shared[72];
- __shared__ float kernel_shared[3072];
+extern "C" __global__ void __launch_bounds__(392) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[2];
+ __shared__ float pad_temp_shared[252];
+ __shared__ float kernel_shared[192];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = ((((1 <= (((int)blockIdx.x) % 7)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73728)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147456)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221184)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294912)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368640)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442368)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516096)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3))];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = (((1 <= (((int)threadIdx.x) % 9)) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 1)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73731)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147459)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221187)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294915)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368643)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442371)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516099)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 3)];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 3)];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- __syncthreads();
- if (((int)threadIdx.x) < 72) {
- pad_temp_shared[((int)threadIdx.x)] = (((((((int)blockIdx.x) % 7) < 6) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 9) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) + 6)] : 0.000000e+00f);
+ for (int rc_outer_outer = 0; rc_outer_outer < 128; ++rc_outer_outer) {
+ for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
+ __syncthreads();
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 <= ((((((int)threadIdx.x) * 4) % 63) / 9) + ry_outer_outer)) && (((((((int)threadIdx.x) * 4) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= ((((int)threadIdx.x) * 4) % 9))) && (((((int)threadIdx.x) * 4) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + (((((int)threadIdx.x) * 4) / 9) * 7)) + (ry_outer_outer * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 1) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 1) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 1) % 9))) && ((((((int)threadIdx.x) * 4) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.00 [...]
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 2) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 2) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 2) % 9))) && ((((((int)threadIdx.x) * 4) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.00 [...]
+ }
+ if (((int)threadIdx.x) < 63) {
+ pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 <= (((((((int)threadIdx.x) * 4) + 3) % 63) / 9) + ry_outer_outer)) && ((((((((int)threadIdx.x) * 4) + 3) % 63) / 9) + ry_outer_outer) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 3) % 9))) && ((((((int)threadIdx.x) * 4) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 196) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 7)) + (ry_outer_outer * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.00 [...]
+ }
+ if (((int)threadIdx.x) < 192) {
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((((int)blockIdx.x) * 73728) + ((((int)threadIdx.x) / 12) * 4608)) + (rc_outer_outer * 36)) + (((((int)threadIdx.x) % 12) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
+ }
+ __syncthreads();
+ for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7))] * kernel_shared[(((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6))]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7))] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 96)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 3)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 63)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 99)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 97)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 64)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 4)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 64)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 100)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 98)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 65)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 5)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 126) + (((((int)threadIdx.x) % 49) / 7) * 9)) + (((int)threadIdx.x) % 7)) + 65)] * kernel_shared[((((((int)threadIdx.x) / 49) * 12) + (rc_outer_inner * 6)) + 101)]));
+ }
}
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 384)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 73734)];
- kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 768)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 147462)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 221190)];
- kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 294918)];
- kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 368646)];
- kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 442374)];
- kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (((int)threadIdx.x) % 3)) + 516102)];
- kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 2) % 3)) + 6)];
- kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + ((((int)threadIdx.x) + 1) % 3)) + 6)];
- __syncthreads();
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 24)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 3)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 1)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 4)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 24) + 2)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 24) + 5)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 6)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 9)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 7)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 10)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 24) + 8)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 24) + 11)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 12)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 15)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 13)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 16)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 24) + 14)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 24) + 17)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 18)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 21)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 19)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 22)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 24) + 20)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 24) + 23)]));
- }
- for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
- compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
}
+ compute[((((int)blockIdx.x) * 784) + ((int)threadIdx.x))] = max((conv2d_nchw[0] + bias[((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 49))]), 0.000000e+00f);
+ compute[(((((int)blockIdx.x) * 784) + ((int)threadIdx.x)) + 392)] = max((conv2d_nchw[1] + bias[(((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 49)) + 8)]), 0.000000e+00f);
}
</pre></div>
</div>
@@ -1964,7 +765,7 @@ 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> ( 5 minutes 35.421 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes 44.391 seconds)</p>
<div class="sphx-glr-footer 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 sphx-glr-download-python 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 e470a2d730..47004fb766 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -916,7 +916,7 @@ 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)
- 7.9023 7.8957 7.9158 7.8954 0.0096
+ 7.8511 7.8500 7.8552 7.8482 0.0030
</pre></div>
</div>
</div>
@@ -938,7 +938,7 @@ 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 1.117 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 6.185 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autoscheduler-tune-network-cuda-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/eafe360d52540634c9eea0fa89e804bd/tune_network_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_network_cuda.py</span></code></a></p>
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 c8b09798e0..8898c142f4 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -935,7 +935,7 @@ 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)
- 754.1883 754.1875 754.8836 753.4939 0.5674
+ 761.4761 760.0342 764.5453 759.8487 2.1716
</pre></div>
</div>
</div>
@@ -957,7 +957,7 @@ 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 30.882 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 40.523 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autoscheduler-tune-network-x86-py">
<div class="sphx-glr-download sphx-glr-download-python 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 131f8257dd..5029b7f00a 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -633,102 +633,78 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [128, 512], []),
compute: Buffer(compute_2: Pointer(float32), float32, [128, 512], [])}
buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute} {
- for (i0.outer: int32, 0, 8) "parallel" {
- allocate(compute_3: Pointer(global float32), float32, [256]), storage_scope = global;
- for (i1.outer: int32, 0, 32) {
- for (i.inner.init: int32, 0, 16) {
- let cse_var_1: int32 = (i.inner.init*16)
- {
- compute_4: Buffer(compute_3, float32, [256], [])[cse_var_1] = 0f32
- compute_4[(cse_var_1 + 1)] = 0f32
- compute_4[(cse_var_1 + 2)] = 0f32
- compute_4[(cse_var_1 + 3)] = 0f32
- compute_4[(cse_var_1 + 4)] = 0f32
- compute_4[(cse_var_1 + 5)] = 0f32
- compute_4[(cse_var_1 + 6)] = 0f32
- compute_4[(cse_var_1 + 7)] = 0f32
- compute_4[(cse_var_1 + 8)] = 0f32
- compute_4[(cse_var_1 + 9)] = 0f32
- compute_4[(cse_var_1 + 10)] = 0f32
- compute_4[(cse_var_1 + 11)] = 0f32
- compute_4[(cse_var_1 + 12)] = 0f32
- compute_4[(cse_var_1 + 13)] = 0f32
- compute_4[(cse_var_1 + 14)] = 0f32
- compute_4[(cse_var_1 + 15)] = 0f32
- }
- }
- for (elem_idx: int32, 0, (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(i1.outer + 1)] - placeholder_15[i1.outer])) {
- for (i.inner: int32, 0, 16) {
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_2: int32 = (i.inner*16)
- compute_4[cse_var_2] = (compute_4[cse_var_2] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[i1.outer]*16) + (elem_idx*16))]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(((i0.outer*4096) + (i.inner*256)) + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_3: int32 = ((i.inner*16) + 1)
- compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 1)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_4: int32 = ((i.inner*16) + 2)
- compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 2)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_5: int32 = ((i.inner*16) + 3)
- compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 3)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_6: int32 = ((i.inner*16) + 4)
- compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 4)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_7: int32 = ((i.inner*16) + 5)
- compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 5)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_8: int32 = ((i.inner*16) + 6)
- compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 6)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_9: int32 = ((i.inner*16) + 7)
- compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 7)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_10: int32 = ((i.inner*16) + 8)
- compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 8)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_11: int32 = ((i.inner*16) + 9)
- compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 9)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_12: int32 = ((i.inner*16) + 10)
- compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 10)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_13: int32 = ((i.inner*16) + 11)
- compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 11)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_14: int32 = ((i.inner*16) + 12)
- compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 12)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_15: int32 = ((i.inner*16) + 13)
- compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 13)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
- }
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_16: int32 = ((i.inner*16) + 14)
- compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 14)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
+ for (i0.outer.i1.outer.fused: int32, 0, 16) "parallel" {
+ allocate(compute_3: Pointer(global float32), float32, [4096]), storage_scope = global {
+ for (i.outer.inner: int32, 0, 4) {
+ for (nb_j.inner: int32, 0, 2) {
+ for (i.inner.init: int32, 0, 32) {
+ let cse_var_1: int32 = (((i.outer.inner*1024) + (i.inner.init*32)) + (nb_j.inner*16))
+ {
+ compute_4: Buffer(compute_3, float32, [4096], [])[cse_var_1] = 0f32
+ compute_4[(cse_var_1 + 1)] = 0f32
+ compute_4[(cse_var_1 + 2)] = 0f32
+ compute_4[(cse_var_1 + 3)] = 0f32
+ compute_4[(cse_var_1 + 4)] = 0f32
+ compute_4[(cse_var_1 + 5)] = 0f32
+ compute_4[(cse_var_1 + 6)] = 0f32
+ compute_4[(cse_var_1 + 7)] = 0f32
+ compute_4[(cse_var_1 + 8)] = 0f32
+ compute_4[(cse_var_1 + 9)] = 0f32
+ compute_4[(cse_var_1 + 10)] = 0f32
+ compute_4[(cse_var_1 + 11)] = 0f32
+ compute_4[(cse_var_1 + 12)] = 0f32
+ compute_4[(cse_var_1 + 13)] = 0f32
+ compute_4[(cse_var_1 + 14)] = 0f32
+ compute_4[(cse_var_1 + 15)] = 0f32
+ }
}
- if @tir.likely((elem_idx < (placeholder_15[(i1.outer + 1)] - placeholder_15[i1.outer])), dtype=bool) {
- let cse_var_17: int32 = ((i.inner*16) + 15)
- compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[i1.outer]*16) + (elem_idx*16)) + 15)]*max(placeholder_17[(((i0.outer*4096) + (i.inner*256)) + placeholder_18[(placeholder_15[i1.outer] + elem_idx)])], 0f32)))
+ for (elem_idx: int32, 0, let cse_var_2: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_2 + 1)] - placeholder_15[cse_var_2])) {
+ for (i.inner: int32, 0, 32) {
+ let cse_var_21: int32 = (elem_idx*16)
+ let cse_var_20: int32 = ((i0.outer.i1.outer.fused*2) + nb_j.inner)
+ let cse_var_19: int32 = ((i.outer.inner*8192) + (i.inner*256))
+ let cse_var_18: int32 = (((i.outer.inner*1024) + (i.inner*32)) + (nb_j.inner*16))
+ let cse_var_17: int32 = (cse_var_18 + 9)
+ let cse_var_16: int32 = (cse_var_18 + 8)
+ let cse_var_15: int32 = (cse_var_18 + 7)
+ let cse_var_14: int32 = (cse_var_18 + 6)
+ let cse_var_13: int32 = (cse_var_18 + 5)
+ let cse_var_12: int32 = (cse_var_18 + 4)
+ let cse_var_11: int32 = (cse_var_18 + 3)
+ let cse_var_10: int32 = (cse_var_18 + 2)
+ let cse_var_9: int32 = (cse_var_18 + 15)
+ let cse_var_8: int32 = (cse_var_18 + 14)
+ let cse_var_7: int32 = (cse_var_18 + 13)
+ let cse_var_6: int32 = (cse_var_18 + 12)
+ let cse_var_5: int32 = (cse_var_18 + 11)
+ let cse_var_4: int32 = (cse_var_18 + 10)
+ let cse_var_3: int32 = (cse_var_18 + 1)
+ {
+ compute_4[cse_var_18] = (compute_4[cse_var_18] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[cse_var_20]*16) + cse_var_21)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(cse_var_19 + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 1)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 2)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 3)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 4)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 5)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 6)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 7)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 8)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 9)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 10)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 11)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 12)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 13)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 14)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 15)]*max(placeholder_17[(cse_var_19 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
+ }
+ }
}
}
}
- for (i0.inner: int32, 0, 16) {
- for (i1.inner: int32, 0, 16) {
- let cse_var_18: int32 = ((((i0.outer*8192) + (i0.inner*512)) + (i1.outer*16)) + i1.inner)
- compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_18] = max((compute_4[((i0.inner*16) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_18]), 0f32)
+ for (i0.inner: int32, 0, 128) {
+ for (i1.inner: int32, 0, 32) {
+ let cse_var_22: int32 = (((i0.inner*512) + (i0.outer.i1.outer.fused*32)) + i1.inner)
+ compute_5: Buffer(compute_2, float32, [65536], [])[cse_var_22] = max((compute_4[((i0.inner*32) + i1.inner)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[cse_var_22]), 0f32)
}
}
}
@@ -767,7 +743,7 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
<span class="p">)</span>
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 2.200 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.804 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 6e9e4fc801..a0f74e68f1 100644
--- a/docs/how_to/tune_with_autotvm/sg_execution_times.html
+++ b/docs/how_to/tune_with_autotvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>00:33.080</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:35.819</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,7 +349,7 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:33.046</p></td>
+<td><p>00:35.785</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
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 f840ec9896..b5c22f1c57 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -690,7 +690,7 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 256, 2, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5246138
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,666420
No: 2 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -813,10 +813,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 1, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,837597
-No: 3 GFLOPS: 1.84/1.84 result: MeasureResult(costs=(0.1259390105,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.4146933555603027, timestamp=1673620329.442073) [('tile_f', [-1, 64, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1228550
-No: 4 GFLOPS: 28.01/28.01 result: MeasureResult(costs=(0.008264531615384615,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9725844860076904, timestamp=1673620331.1912718) [('tile_f', [-1, 1, 4, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,493609
-No: 5 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9432354
+No: 3 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -938,8 +936,9 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7478219
-No: 6 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3730260
+No: 4 GFLOPS: 3.76/3.76 result: MeasureResult(costs=(0.06163596725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.509656190872192, timestamp=1673672256.9632695) [('tile_f', [-1, 1, 1, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4107784
+No: 5 GFLOPS: 0.00/3.76 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1061,9 +1060,9 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9627989
-No: 7 GFLOPS: 8.42/28.01 result: MeasureResult(costs=(0.02750208075,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4386982917785645, timestamp=1673620333.7890215) [('tile_f', [-1, 2, 2, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3256671
-No: 8 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6703134
+No: 6 GFLOPS: 86.57/86.57 result: MeasureResult(costs=(0.0026740413421052632,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.388312816619873, timestamp=1673672259.522572) [('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1393554
+No: 7 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1185,9 +1184,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 2, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9208051
-No: 9 GFLOPS: 2.10/28.01 result: MeasureResult(costs=(0.11018129725,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.1899983882904053, timestamp=1673620336.2825646) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 256, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8546562
-No: 10 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2058261
+No: 8 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1309,8 +1307,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10028778
-No: 11 GFLOPS: 0.00/28.01 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10227168
+No: 9 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1432,9 +1430,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 2, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,407833
-No: 12 GFLOPS: 200.73/200.73 result: MeasureResult(costs=(0.0011533035517241378,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1821565628051758, timestamp=1673620337.0173085) [('tile_f', [-1, 2, 32, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3507991
-No: 13 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 8, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2476476
+No: 10 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1556,8 +1553,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,375480
-No: 14 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 1, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1183142
+No: 11 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1679,8 +1676,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 64, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3174867
-No: 15 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 32, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3651501
+No: 12 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1802,8 +1799,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 4, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2115377
-No: 16 GFLOPS: 0.00/200.73 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 1, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7328306
+No: 13 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1925,10 +1922,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 128, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7160658
-No: 17 GFLOPS: 239.04/239.04 result: MeasureResult(costs=(0.0009684793679245284,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2931921482086182, timestamp=1673620338.6904929) [('tile_f', [-1, 4, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3560522
-No: 18 GFLOPS: 0.95/239.04 result: MeasureResult(costs=(0.24484147625,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.071900367736816, timestamp=1673620342.299537) [('tile_f', [-1, 1, 1, 256]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,812236
-No: 19 GFLOPS: 0.00/239.04 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,495707
+No: 14 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -2050,8 +2045,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8625470
-No: 20 GFLOPS: 0.00/239.04 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 256, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9377333
+No: 15 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -2173,7 +2168,395 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 4, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4975596
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5716292
+No: 16 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2222920
+No: 17 GFLOPS: 0.00/86.57 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
+ return self.__get_result()
+ File "/usr/lib/python3.7/concurrent/futures/_base.py", line 384, in __get_result
+ raise self._exception
+ File "/usr/lib/python3.7/concurrent/futures/thread.py", line 57, in run
+ result = self.fn(*self.args, **self.kwargs)
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 432, in <lambda>
+ worker = lambda *args: self._worker_run(*args)
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 401, in _worker_run
+ return proc.recv()
+ File "/workspace/python/tvm/contrib/popen_pool.py", line 309, in recv
+ raise TimeoutError()
+TimeoutError
+
+ [('tile_f', [-1, 1, 2, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9078804
+No: 18 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2920237
+No: 19 GFLOPS: 0.00/86.57 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
+ func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
+ func = build(s, args, target_host=task.target_host, runtime=runtime)
+ File "/workspace/python/tvm/driver/build_module.py", line 227, in build
+ input_mod = lower(inputs, args, name=name, binds=binds)
+ File "/workspace/python/tvm/driver/build_module.py", line 134, in lower
+ return ffi.lower_schedule(inp, args, name, binds, simple_mode)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel
+
+Traceback (most recent call last):
+ 24: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 23: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 22: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 21: operator()
+ at ../include/tvm/runtime/packed_func.h:1730
+ 20: unpack_call<tvm::IRModule, 5, tvm::<lambda(tvm::te::Schedule, const tvm::runtime::Array<tvm::runtime::ObjectRef>&, const tvm::runtime::String&, const tvm::runtime::Map<tvm::te::Tensor, tvm::tir::Buffer>&, bool)> >
+ at ../include/tvm/runtime/packed_func.h:1670
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1630
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1645
+ 13: operator()
+ at ../src/driver/driver_api.cc:395
+ 12: tvm::LowerSchedule(tvm::te::Schedule, tvm::runtime::Array<tvm::runtime::ObjectRef, void> const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::unordered_map<tvm::te::Tensor, tvm::tir::Buffer, std::hash<tvm::te::Tensor>, std::equal_to<tvm::te::Tensor>, std::allocator<std::pair<tvm::te::Tensor const, tvm::tir::Buffer> > > const&, tvm::GlobalVarSupply, bool)
+ at ../src/driver/driver_api.cc:381
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:276
+ 10: tvm::transform::Pass::operator()(tvm::IRModule) const
+ at ../src/ir/transform.cc:258
+ 9: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 8: tvm::transform::SequentialNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:454
+ 7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/ir/transform.cc:274
+ 6: tvm::tir::transform::PrimFuncPassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
+ at ../src/tir/ir/transform.cc:100
+ 5: tvm::runtime::TypedPackedFunc<tvm::tir::PrimFunc (tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext)>::operator()(tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext) const
+ at ../include/tvm/runtime/packed_func.h:1749
+ 4: tvm::tir::PrimFunc tvm::runtime::detail::typed_packed_call_dispatcher<tvm::tir::PrimFunc>::run<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::runtime::PackedFunc const&, tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&)
+ at ../include/tvm/runtime/packed_func.h:1693
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<tvm::tir::PrimFunc, tvm::IRModule, tvm::transform::PassContext>(tvm::tir::PrimFunc&&, tvm::IRModule&&, tvm::transform::PassContext&&) const
+ at ../include/tvm/runtime/packed_func.h:1617
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, 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, 16, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9207968
+No: 20 GFLOPS: 77.94/86.57 result: MeasureResult(costs=(0.002970176054054054,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.8064920902252197, timestamp=1673672270.9395628) [('tile_f', [-1, 2, 1, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3300321
</pre></div>
</div>
<p>Finally we can inspect the best config from log file, check correctness,
@@ -2212,9 +2595,9 @@ and measure running time.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Finish loading 20 records
Best config:
-[('tile_f', [-1, 4, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3560522
+[('tile_f', [-1, 4, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1393554
Finish loading 20 records
-Time cost of this operator: 0.001384
+Time cost of this operator: 0.001601
</pre></div>
</div>
<div class="sphx-glr-footer 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 a7fb08e188..232d9ea377 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -646,10 +646,10 @@ 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 Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 310.0 98.724 (1, 2, 10, 10, 3) 2 1 [310.0]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.021 0.962 (1, 6, 10, 10) 1 1 [3.021]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.985 0.314 (1, 1, 10, 10, 3) 1 1 [0.985]
-Total_time - 314.007 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 312.3 98.731 (1, 2, 10, 10, 3) 2 1 [312.3]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.026 0.957 (1, 6, 10, 10) 1 1 [3.026]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.987 0.312 (1, 1, 10, 10, 3) 1 1 [0.987]
+Total_time - 316.313 - - - - -
</pre></div>
</div>
</div>
@@ -701,10 +701,10 @@ 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 Measurements(us)
--------- --- -------- ------- ----- ------ ------- ----------------
-tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 102.1 97.391 (1, 6, 10, 10, 1) 2 1 [102.1]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.775 1.693 (1, 6, 10, 10) 1 1 [1.775]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.96 0.916 (1, 1, 10, 10, 3) 1 1 [0.96]
-Total_time - 104.835 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 139.2 98.15 (1, 6, 10, 10, 1) 2 1 [139.2]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.781 1.256 (1, 6, 10, 10) 1 1 [1.781]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.842 0.594 (1, 3, 10, 10, 1) 1 1 [0.842]
+Total_time - 141.823 - - - - -
</pre></div>
</div>
<div class="sphx-glr-footer 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_pytorch.html b/docs/how_to/work_with_microtvm/micro_pytorch.html
index 7ae8cd782f..31cee8d678 100644
--- a/docs/how_to/work_with_microtvm/micro_pytorch.html
+++ b/docs/how_to/work_with_microtvm/micro_pytorch.html
@@ -453,7 +453,7 @@ download a cat image and preprocess it to use as the model input.</p>
Downloading: "https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
0%| | 0.00/3.42M [00:00<?, ?B/s]
-100%|##########| 3.42M/3.42M [00:00<00:00, 45.5MB/s]
+100%|##########| 3.42M/3.42M [00:00<00:00, 45.6MB/s]
/workspace/python/tvm/relay/frontend/pytorch_utils.py:47: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
return LooseVersion(torch_ver) > ver
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/setuptools/_distutils/version.py:346: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
@@ -577,7 +577,7 @@ via the host <cite>main.cc`</cite> or if a Zephyr emulated board is selected as
Torch top-1 id: 282, class name: tiger cat
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 2.311 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 11.209 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-work-with-microtvm-micro-pytorch-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/12b9ecc04c41abaa12022061771821d1/micro_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">micro_pytorch.py</span></code></a></p>
diff --git a/docs/how_to/work_with_microtvm/micro_train.html b/docs/how_to/work_with_microtvm/micro_train.html
index e221a466f7..89c1c715dd 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -523,7 +523,7 @@ take about <strong>2 minutes</strong> to download the Stanford Cars, while COCO
<a href="https://docs.python.org/3/library/shutil.html#shutil.move" title="shutil.move" class="sphx-glr-backref-module-shutil sphx-glr-backref-type-py-function"><span class="n">shutil</span><span class="o">.</span><span class="n">move</span></a><span class="p">(</span><span class="sa">f</span><span class="s2">"</span><span class="si">{</span><a href="https://docs.python.org/3/library/stdtypes.html#str" title="builtins.str" class="sphx-glr-backref-module-builtins sphx-glr-backref-typ [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpbg21p4m5/images/random'
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmpv7f1wxl6/images/random'
</pre></div>
</div>
</div>
@@ -583,8 +583,8 @@ objects to other stuff? We can display some examples from our datasets using <co
<span class="n">plt</span><span class="o">.</span><span class="n">axis</span><span class="p">(</span><span class="s2">"off"</span><span class="p">)</span>
</pre></div>
</div>
-<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0]" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>/tmp/tmpbg21p4m5/images/target contains 8144 images
-/tmp/tmpbg21p4m5/images/random contains 5000 images
+<img src="../../_images/sphx_glr_micro_train_001.png" srcset="../../_images/sphx_glr_micro_train_001.png" alt="[0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [1.0, 0.0]" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>/tmp/tmpv7f1wxl6/images/target contains 8144 images
+/tmp/tmpv7f1wxl6/images/random contains 5000 images
</pre></div>
</div>
</div>
@@ -696,13 +696,13 @@ the time on our validation set).</p>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Epoch 1/3
-328/328 - 47s - loss: 0.2198 - accuracy: 0.9223 - val_loss: 0.1192 - val_accuracy: 0.9600 - 47s/epoch - 143ms/step
+328/328 - 47s - loss: 0.2314 - accuracy: 0.9218 - val_loss: 0.1113 - val_accuracy: 0.9573 - 47s/epoch - 144ms/step
Epoch 2/3
-328/328 - 43s - loss: 0.1058 - accuracy: 0.9609 - val_loss: 0.1799 - val_accuracy: 0.9407 - 43s/epoch - 132ms/step
+328/328 - 44s - loss: 0.1005 - accuracy: 0.9632 - val_loss: 0.1100 - val_accuracy: 0.9607 - 44s/epoch - 133ms/step
Epoch 3/3
-328/328 - 43s - loss: 0.0606 - accuracy: 0.9779 - val_loss: 0.0981 - val_accuracy: 0.9687 - 43s/epoch - 132ms/step
+328/328 - 44s - loss: 0.0765 - accuracy: 0.9730 - val_loss: 0.1205 - val_accuracy: 0.9607 - 44s/epoch - 133ms/step
-<keras.callbacks.History object at 0x7fae16c89150>
+<keras.callbacks.History object at 0x7f14b419db50>
</pre></div>
</div>
</div>
@@ -962,7 +962,7 @@ as intended.</p>
<p>From here, we could modify the model to read live images from the camera - we have another
Arduino tutorial for how to do that <a class="reference external" href="https://github.com/guberti/tvm-arduino-demos/tree/master/examples/person_detection">on GitHub</a>. Alternatively, we could also
<a class="reference external" href="https://tvm.apache.org/docs/how_to/work_with_microtvm/micro_autotune.html">use TVM’s autotuning capabilities</a> to dramatically improve the model’s performance.</p>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 53.524 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 28.535 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-work-with-microtvm-micro-train-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/b52cec46baf4f78d6bcd94cbe269c8a6/micro_train.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">micro_train.py</span></code></a></p>
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 179ae28acc..ac767c9c2f 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>06:58.377</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>06:46.352</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,30 +349,30 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_train.html#sphx-glr-how-to-work-with-microtvm-micro-train-py"><span class="std std-ref">Training Vision Models for microTVM on Arduino</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_train.py</span></code>)</p></td>
-<td><p>04:53.524</p></td>
+<td><p>04:28.535</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_pytorch.html#sphx-glr-how-to-work-with-microtvm-micro-pytorch-py"><span class="std std-ref">microTVM PyTorch Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_pytorch.py</span></code>)</p></td>
-<td><p>01:02.311</p></td>
+<td><p>01:11.209</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:50.938</p></td>
+<td><p>00:53.735</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="micro_aot.html#sphx-glr-how-to-work-with-microtvm-micro-aot-py"><span class="std std-ref">microTVM Host-Driven AoT</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_aot.py</span></code>)</p></td>
-<td><p>00:07.791</p></td>
+<td><p>00:08.938</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:03.812</p></td>
+<td><p>00:03.935</p></td>
<td><p>0.0 MB</p></td>
</tr>
-<tr class="row-even"><td><p><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></td>
+<tr class="row-even"><td><p><a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></td>
<td><p>00:00.000</p></td>
<td><p>0.0 MB</p></td>
</tr>
-<tr class="row-odd"><td><p><a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></td>
+<tr class="row-odd"><td><p><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></td>
<td><p>00:00.000</p></td>
<td><p>0.0 MB</p></td>
</tr>
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 d767f02054..8ae11f994e 100644
--- a/docs/how_to/work_with_relay/sg_execution_times.html
+++ b/docs/how_to/work_with_relay/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>00:44.187</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:45.858</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,15 +349,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="using_pipeline_executor.html#sphx-glr-how-to-work-with-relay-using-pipeline-executor-py"><span class="std std-ref">Using Pipeline Executor in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_pipeline_executor.py</span></code>)</p></td>
-<td><p>00:32.432</p></td>
+<td><p>00:33.692</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:10.283</p></td>
+<td><p>00:10.362</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:01.467</p></td>
+<td><p>00:01.797</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
diff --git a/docs/how_to/work_with_schedules/intrin_math.html b/docs/how_to/work_with_schedules/intrin_math.html
index 815ee8198b..ca6d29566e 100644
--- a/docs/how_to/work_with_schedules/intrin_math.html
+++ b/docs/how_to/work_with_schedules/intrin_math.html
@@ -535,7 +535,7 @@ The following example customizes CUDA lowering rule for <code class="code docuti
<a href="../../reference/api/python/ir.html#tvm.ir.register_intrin_lowering" title="tvm.ir.register_intrin_lowering" class="sphx-glr-backref-module-tvm-ir sphx-glr-backref-type-py-function"><span class="n">register_intrin_lowering</span></a><span class="p">(</span><span class="s2">"tir.exp"</span><span class="p">,</span> <span class="n">target</span><span class="o">=</span><span class="s2">"cuda"</span><span class="p">,</span> <span class="n">f</span><span class="o">= [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7fae1663b950>
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7f14b46ccb00>
</pre></div>
</div>
<p>Register the rule to TVM with override option to override existing rule.
diff --git a/docs/how_to/work_with_schedules/reduction.html b/docs/how_to/work_with_schedules/reduction.html
index 6e406629ca..fc7bb4445f 100644
--- a/docs/how_to/work_with_schedules/reduction.html
+++ b/docs/how_to/work_with_schedules/reduction.html
@@ -557,7 +557,7 @@ the first axis of reduced result of B.f</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">print</span><span class="p">(</span><a href="../../reference/api/python/te.html#tvm.te.Schedule" title="tvm.te.Schedule" class="sphx-glr-backref-module-tvm-te sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">s</span></a><span class="p">[</span><a href="../../reference/api/python/te.html#tvm.te.Tensor" title="tvm.te.Tensor" class="sphx-glr-backref-module-tvm-te sph [...]
</pre></div>
</div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>[reduce(combiner=comm_reducer(result=[(x + y)], lhs=[x], rhs=[y], identity_element=[0f]), source=[B.rf[k.inner.v, ax0]], init=[], axis=[iter_var(k.inner.v, range(min=0, ext=16))], where=(bool)1, value_index=0)]
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>[T.reduce(T.comm_reducer(lambda x, y: x + y, [T.float32(0)]), source=[B.rf[k_inner_v, ax0]], init=[], axis=[T.iter_var(k_inner_v, T.Range(0, 16), "CommReduce", "")], condition=True, value_index=0)]
</pre></div>
</div>
</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 fda26be4ad..9c04cc94c5 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -340,7 +340,7 @@
<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>00:05.106</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:07.535</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,35 +349,35 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:02.513</p></td>
+<td><p>00:04.995</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:01.187</p></td>
+<td><p>00:01.136</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:00.605</p></td>
+<td><p>00:00.603</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:00.589</p></td>
+<td><p>00:00.573</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:00.112</p></td>
+<td><p>00:00.120</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:00.049</p></td>
+<td><p>00:00.051</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><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></td>
-<td><p>00:00.029</p></td>
+<td><p>00:00.033</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><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></td>
-<td><p>00:00.023</p></td>
+<td><p>00:00.024</p></td>
<td><p>0.0 MB</p></td>
</tr>
</tbody>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index 9470c42d54..3567bc658d 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -587,7 +587,7 @@ The importing needs to happen before the tensorized GEMV being executed.</p>
B: Buffer(B_2: Pointer(float32), float32, [512, 64], []),
C: Buffer(C_2: Pointer(float32), float32, [1024, 512], [])}
buffer_map = {A_1: A, B_1: B, C_1: C} {
- attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpfqspl5_8/input0.cc'\nsource_filename = \"/tmp/tmpfqspl5_8/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/tmpls10z2ft/input0.cc'\nsource_filename = \"/tmp/tmpls10z2ft/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/classes.html b/docs/reference/api/doxygen/classes.html
index 0beecdc433..3fec51a220 100644
--- a/docs/reference/api/doxygen/classes.html
+++ b/docs/reference/api/doxygen/classes.html
@@ -103,8 +103,8 @@ $(function() {
<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_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_1arith_1_1IterSumExprNode.html [...]
<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_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_1tir_1_1IterVar.html">IterVar</a> (<a class="el" href="na [...]
<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="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="classtvm_1_1te_1_1IterVarAttr.html">IterVarAttr</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_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_1te_1_1IterVarAttrNode.html">IterVarAttrNode</a> (<a class="el" href="namespacetvm_1_ [...]
-<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_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_1IterVarNode.html">IterVarNode</a> (<a class= [...]
+<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_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_1te_1_1IterVarAttrNode.html">IterVarAttrNode</a> (<a class="el" href="namespacetvm_1_ [...]
+<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_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_1IterVarNode.html">IterVarNode</a> (<a class= [...]
<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_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_1te_1_1IterVarRelation.html">IterVar [...]
<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_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_1te_1_1IterVarRelationNode.html">IterVarRelationNode</a> (<a clas [...]
<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_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 rowspan="2" valig [...]
@@ -141,16 +141,16 @@ $(function() {
<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="classtvm_1_1script_1_1printer_1_1ListDoc.html">ListDoc</a> (<a class="el" href="namespacetvm_1_1script_1_1printer.html">tvm::script::printer</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1te_1_1TensorComputeOp.html">TensorComputeOp</a> (<a class="el" href= [...]
<tr><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AttrFrame.html">AttrFrame</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</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_1script_1_1printer_1_ [...]
<tr><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AttrFrameNode.html">AttrFrameNode</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1topi_1_1EinsumEquation.html">EinsumEquation</a> (<a class="el" href="namespacetvm_1_1topi.html">tvm::topi</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_ [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttributeAccessPath.html">AttributeAccessPath</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrame.html">ElseFrame</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1printer_1_1LiteralDo [...]
-<tr><td valign="top"><a class="el" href="classtvm_1_1AttributeAccessPathNode.html">AttributeAccessPathNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrameNode.html">ElseFrameNode</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Load.ht [...]
-<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_1EnvFunc.html">EnvFunc</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</a>)   [...]
-<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_1EnvFuncNode.html">EnvFuncNode</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> (<a class="el" href="namespacetvm_1_1aut [...]
-<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_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_1auto__scheduler_1_1LocalBuilderNode.html">LocalBuilderNode</a> (<a class="el" href [...]
-<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_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_1auto__scheduler_1_1LocalRunner.html">LocalRunner</a> (<a class="el" href="namespacetvm_1_1a [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttributeAccessPath.html">AttributeAccessPath</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrame.html">ElseFrame</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1printer_1_1LiteralDo [...]
+<tr><td valign="top"><a class="el" href="classtvm_1_1AttributeAccessPathNode.html">AttributeAccessPathNode</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrameNode.html">ElseFrameNode</a> (<a class="el" href="namespacetvm_1_1script_1_1ir__builder_1_1tir.html">tvm::script::ir_builder::tir</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1tir_1_1Load.ht [...]
+<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_1EnvFunc.html">EnvFunc</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</a>)   [...]
+<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_1EnvFuncNode.html">EnvFuncNode</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> (<a class="el" href="namespacetvm_1_1aut [...]
+<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_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_1auto__scheduler_1_1LocalBuilderNode.html">LocalBuilderNode</a> (<a class="el" href [...]
+<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_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_1auto__scheduler_1_1LocalRunner.html">LocalRunner</a> (<a class="el" href="namespacetvm_1_1a [...]
<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="structtvm_1_1ErrorBuilder.html">ErrorBuilder</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="namesp [...]
<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_1ErrorReporter.html">ErrorReporter</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_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_1Evaluate.html">Evaluate</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</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" [...]
-<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_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_1LRNAttrs.html">LRNAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">tvm::relay</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="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="classtvm_1_1tir_1_1LoopRVNode.html">LoopRVNode</a> (<a class="el" href="namespacetvm_1_1tir.html" [...]
+<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_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_1LRNAttrs.html">LRNAttrs</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_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_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_1AttrRegistryMapContainerMap.html">AttrRegistryMapContainerMap</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_1tir_1_1LTNode.html">LTNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tv [...]
<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_1relay_1_1ExecutorNode.html">ExecutorNode</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">  m  </div></td> [...]
@@ -171,8 +171,8 @@ $(function() {
<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="classtvm_1_1tir_1_1MaxNode.html">MaxNode</a> (<a class="el" href="namespacetvm_1_1tir.html">tvm::tir</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1RefValueObj.html">RefValueObj</a> (<a class="el" href="namespacetvm_1_1relay. [...]
<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="structtvm_1_1relay_1_1MaxPool1DAttrs.html">MaxPool1DAttrs</a> (<a class="el" href="namespacetvm_1_1relay.html">t [...]
<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_1script_1_1printer_1_1ExprStmtDoc.html">ExprStmtDoc</a> (<a class="el" href="namespacetvm_1_1script_1_1printer.html">tvm::script::printer</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool2DAttrs.html">MaxPool2DAttr [...]
-<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_1script_1_1printer_1_1ExprStmtDocNode.html">ExprStmtDocNode</a> (<a class="el" href="namespacetvm_1_1script_1_1printer.html">tvm::script::printer</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool3DAttrs.html">MaxPool3DAttrs</a> (<a class="el" href= [...]
-<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_1meta__schedule_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" href="namespacetvm_1_1meta__schedul [...]
+<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_1script_1_1printer_1_1ExprStmtDocNode.html">ExprStmtDocNode</a> (<a class="el" href="namespacetvm_1_1script_1_1printer.html">tvm::script::printer</a>)   </td><td valign="top"><a class="el" href="structtvm_1_1relay_1_1MaxPool3DAttrs.html">MaxPool3DAttrs</a> (<a class="el" href= [...]
+<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_1meta__schedule_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" href="namespacetvm_1_1meta__schedul [...]
<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_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_1auto__scheduler_1_1MeasureCallback.html">MeasureCallback</a> (<a class="el" href="namespacetvm_1_1auto__schedule [...]
<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_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_1meta__schedule_1_1MeasureCallbackNode.html">MeasureCallbackNode</a> (<a class="el" href="namespacetvm_1_1meta__schedule [...]
<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_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_1auto__scheduler_1_1MeasureCallbackNode.html">MeasureCallbackNode</a> (<a class="el" href="namespacetvm_1_1a [...]
@@ -307,8 +307,8 @@ $(function() {
<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_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_1instrument_1_1PassInstrumentNode.html">PassInstrumentNode</a> (<a class="el" hre [...]
<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_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_1transform_1_1PassNode.html">PassNode</a> (<a class="el" href="namespa [...]
<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_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_1Pattern.html">Pattern</a> (<a class="el" href="na [...]
-<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_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_1PatternConstructor.html">PatternCons [...]
-<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_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_1PatternConstructor [...]
+<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_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_1PatternConstructor.html">PatternCons [...]
+<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_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_1PatternConstructor [...]
<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_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_1PatternFunctor.html">PatternFunctor</a> [...]
<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_1Integer.html">Integer</a> (<a class="el" href="namespacetvm.html">tvm</a>)   </td><td valign="top"><a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html">PatternFunctor [...]
</td></tr>
diff --git a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef.html b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef.html
index 8ed241058e..27fe2c04a2 100644
--- a/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef.html
+++ b/docs/reference/api/doxygen/classtvm_1_1runtime_1_1ObjectRef.html
@@ -81,7 +81,7 @@ $(function() {
<p><code>#include <<a class="el" href="object_8h_source.html">object.h</a>></code></p>
-<p>Inherited by <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< Range ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< Region ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< tvm::arith::IterSplitExpr ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< tvm::arith::IterSumExpr ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tv [...]
+<p>Inherited by <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< Range ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< Region ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< tvm::arith::IterSplitExpr ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tvm::runtime::Array< tvm::arith::IterSumExpr ></a>, <a class="el" href="classtvm_1_1runtime_1_1Array.html">tv [...]
<div class="dynheader">
Collaboration diagram for tvm::runtime::ObjectRef:</div>
<div class="dyncontent">
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 98792aa0ad..c2560dc694 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"/>
-<!-- Node482 -->
+<!-- Node481 -->
<g id="node1" class="node">
-<title>Node482</title>
+<title>Node481</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>
-<!-- Node483 -->
+<!-- Node482 -->
<g id="node2" class="node">
-<title>Node483</title>
+<title>Node482</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>
-<!-- Node483->Node482 -->
+<!-- Node482->Node481 -->
<g id="edge1" class="edge">
-<title>Node483->Node482</title>
+<title>Node482->Node481</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_1script_1_1printer_1_1FrameNode.html b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode.html
index 8a72409c97..f3b67e9100 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode.html
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode.html
@@ -82,7 +82,7 @@ Inheritance diagram for tvm::script::printer::FrameNode:</div>
<div class="dynheader">
Collaboration diagram for tvm::script::printer::FrameNode:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1script_1_1printer_1_1FrameNode__coll__graph.svg" width="1219" height="1988"><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_1script_1_1printer_1_1FrameNode__coll__graph.svg" width="1020" height="1988"><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">
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode__coll__graph.svg b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode__coll__graph.svg
index 968f6d5abe..4831ff13fc 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode__coll__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1FrameNode__coll__graph.svg
@@ -4,369 +4,331 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: tvm::script::printer::FrameNode Pages: 1 -->
-<svg width="914pt" height="1491pt"
- viewBox="0.00 0.00 914.12 1491.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<svg width="765pt" height="1491pt"
+ viewBox="0.00 0.00 765.18 1491.00" 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 1487)">
<title>tvm::script::printer::FrameNode</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-1487 910.1182,-1487 910.1182,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-1487 761.1816,-1487 761.1816,4 -4,4"/>
<!-- Node2 -->
<g id="node1" class="node">
<title>Node2</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="55.6182,-.5 55.6182,-145.5 262.6182,-145.5 262.6182,-.5 55.6182,-.5"/>
-<text text-anchor="start" x="63.6182" y="-133.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
-<text text-anchor="middle" x="159.1182" y="-122.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::FrameNode</text>
-<polyline fill="none" stroke="#000000" points="55.6182,-115.5 262.6182,-115.5 "/>
-<text text-anchor="start" x="63.6182" y="-103.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ callbacks</text>
-<text text-anchor="start" x="63.6182" y="-92.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<polyline fill="none" stroke="#000000" points="55.6182,-85.5 262.6182,-85.5 "/>
-<text text-anchor="start" x="63.6182" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
-<text text-anchor="start" x="63.6182" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
-<text text-anchor="start" x="63.6182" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~FrameNode()</text>
-<text text-anchor="start" x="63.6182" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AddExitCallback()</text>
-<text text-anchor="start" x="63.6182" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AddDispatchToken()</text>
-<text text-anchor="start" x="63.6182" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ EnterWithScope()</text>
-<text text-anchor="start" x="63.6182" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ExitWithScope()</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="53.6816,-.5 53.6816,-145.5 260.6816,-145.5 260.6816,-.5 53.6816,-.5"/>
+<text text-anchor="start" x="61.6816" y="-133.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
+<text text-anchor="middle" x="157.1816" y="-122.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::FrameNode</text>
+<polyline fill="none" stroke="#000000" points="53.6816,-115.5 260.6816,-115.5 "/>
+<text text-anchor="start" x="61.6816" y="-103.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ callbacks</text>
+<text text-anchor="start" x="61.6816" y="-92.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="53.6816,-85.5 260.6816,-85.5 "/>
+<text text-anchor="start" x="61.6816" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
+<text text-anchor="start" x="61.6816" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_BASE_OBJECT_INFO()</text>
+<text text-anchor="start" x="61.6816" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~FrameNode()</text>
+<text text-anchor="start" x="61.6816" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AddExitCallback()</text>
+<text text-anchor="start" x="61.6816" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AddDispatchToken()</text>
+<text text-anchor="start" x="61.6816" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ EnterWithScope()</text>
+<text text-anchor="start" x="61.6816" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ExitWithScope()</text>
</g>
<!-- Node3 -->
<g id="node2" class="node">
<title>Node3</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="67.6182,-609.5 67.6182,-996.5 250.6182,-996.5 250.6182,-609.5 67.6182,-609.5"/>
-<text text-anchor="middle" x="159.1182" y="-984.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
-<polyline fill="none" stroke="#000000" points="67.6182,-977.5 250.6182,-977.5 "/>
-<text text-anchor="start" x="75.6182" y="-965.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<text text-anchor="start" x="75.6182" y="-954.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_final</text>
-<text text-anchor="start" x="75.6182" y="-943.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots</text>
-<text text-anchor="start" x="75.6182" y="-932.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots_can</text>
-<text text-anchor="start" x="75.6182" y="-921.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_overflow</text>
-<text text-anchor="start" x="75.6182" y="-910.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_visit</text>
-<text text-anchor="start" x="75.6182" y="-899.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_attrs</text>
-<text text-anchor="start" x="75.6182" y="-888.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_sequal</text>
-<text text-anchor="start" x="75.6182" y="-877.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="75.6182" y="-866.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_shash</text>
-<text text-anchor="start" x="75.6182" y="-855.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="75.6182" y="-844.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_index</text>
-<text text-anchor="start" x="75.6182" y="-833.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># type_index_</text>
-<text text-anchor="start" x="75.6182" y="-822.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># ref_counter_</text>
-<polyline fill="none" stroke="#000000" points="67.6182,-815.5 250.6182,-815.5 "/>
-<text text-anchor="start" x="75.6182" y="-803.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_index()</text>
-<text text-anchor="start" x="75.6182" y="-792.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKey()</text>
-<text text-anchor="start" x="75.6182" y="-781.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKeyHash()</text>
-<text text-anchor="start" x="75.6182" y="-770.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsInstance()</text>
-<text text-anchor="start" x="75.6182" y="-759.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
-<text text-anchor="start" x="75.6182" y="-748.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="75.6182" y="-737.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="75.6182" y="-726.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="75.6182" y="-715.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="75.6182" y="-704.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="75.6182" y="-693.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2Key()</text>
-<text text-anchor="start" x="75.6182" y="-682.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2KeyHash()</text>
-<text text-anchor="start" x="75.6182" y="-671.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeKey2Index()</text>
-<text text-anchor="start" x="75.6182" y="-660.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _GetOrAllocRuntimeTypeIndex()</text>
-<text text-anchor="start" x="75.6182" y="-649.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RuntimeTypeIndex()</text>
-<text text-anchor="start" x="75.6182" y="-638.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># IncRef()</text>
-<text text-anchor="start" x="75.6182" y="-627.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DecRef()</text>
-<text text-anchor="start" x="75.6182" y="-616.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetOrAllocRuntimeTypeIndex()</text>
+<polygon fill="#ffffff" stroke="#000000" points="65.6816,-609.5 65.6816,-996.5 248.6816,-996.5 248.6816,-609.5 65.6816,-609.5"/>
+<text text-anchor="middle" x="157.1816" y="-984.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
+<polyline fill="none" stroke="#000000" points="65.6816,-977.5 248.6816,-977.5 "/>
+<text text-anchor="start" x="73.6816" y="-965.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<text text-anchor="start" x="73.6816" y="-954.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_final</text>
+<text text-anchor="start" x="73.6816" y="-943.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots</text>
+<text text-anchor="start" x="73.6816" y="-932.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots_can</text>
+<text text-anchor="start" x="73.6816" y="-921.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_overflow</text>
+<text text-anchor="start" x="73.6816" y="-910.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_visit</text>
+<text text-anchor="start" x="73.6816" y="-899.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_attrs</text>
+<text text-anchor="start" x="73.6816" y="-888.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_sequal</text>
+<text text-anchor="start" x="73.6816" y="-877.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
+<text text-anchor="start" x="73.6816" y="-866.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_shash</text>
+<text text-anchor="start" x="73.6816" y="-855.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
+<text text-anchor="start" x="73.6816" y="-844.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_index</text>
+<text text-anchor="start" x="73.6816" y="-833.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># type_index_</text>
+<text text-anchor="start" x="73.6816" y="-822.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># ref_counter_</text>
+<polyline fill="none" stroke="#000000" points="65.6816,-815.5 248.6816,-815.5 "/>
+<text text-anchor="start" x="73.6816" y="-803.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_index()</text>
+<text text-anchor="start" x="73.6816" y="-792.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKey()</text>
+<text text-anchor="start" x="73.6816" y="-781.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKeyHash()</text>
+<text text-anchor="start" x="73.6816" y="-770.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsInstance()</text>
+<text text-anchor="start" x="73.6816" y="-759.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
+<text text-anchor="start" x="73.6816" y="-748.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
+<text text-anchor="start" x="73.6816" y="-737.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
+<text text-anchor="start" x="73.6816" y="-726.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
+<text text-anchor="start" x="73.6816" y="-715.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="73.6816" y="-704.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="73.6816" y="-693.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2Key()</text>
+<text text-anchor="start" x="73.6816" y="-682.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2KeyHash()</text>
+<text text-anchor="start" x="73.6816" y="-671.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeKey2Index()</text>
+<text text-anchor="start" x="73.6816" y="-660.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _GetOrAllocRuntimeTypeIndex()</text>
+<text text-anchor="start" x="73.6816" y="-649.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RuntimeTypeIndex()</text>
+<text text-anchor="start" x="73.6816" y="-638.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># IncRef()</text>
+<text text-anchor="start" x="73.6816" y="-627.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DecRef()</text>
+<text text-anchor="start" x="73.6816" y="-616.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetOrAllocRuntimeTypeIndex()</text>
</a>
</g>
</g>
<!-- Node3->Node2 -->
<g id="edge1" class="edge">
<title>Node3->Node2</title>
-<path fill="none" stroke="#191970" d="M159.1182,-598.9387C159.1182,-447.6873 159.1182,-250.3337 159.1182,-145.6351"/>
-<polygon fill="none" stroke="#191970" points="155.6183,-599.2899 159.1182,-609.29 162.6183,-599.29 155.6183,-599.2899"/>
+<path fill="none" stroke="#191970" d="M157.1816,-598.9387C157.1816,-447.6873 157.1816,-250.3337 157.1816,-145.6351"/>
+<polygon fill="none" stroke="#191970" points="153.6817,-599.2899 157.1816,-609.29 160.6817,-599.29 153.6817,-599.2899"/>
</g>
<!-- Node3->Node3 -->
<g id="edge2" class="edge">
<title>Node3->Node3</title>
-<path fill="none" stroke="#404040" d="M250.9807,-836.9248C261.6664,-830.6637 268.6182,-819.3555 268.6182,-803 268.6182,-792.0112 265.48,-783.3007 260.1797,-776.8687"/>
-<polygon fill="none" stroke="#404040" points="260.1366,-776.8322 252.973,-776.0056 250.9807,-769.0752 258.1443,-769.9017 260.1366,-776.8322"/>
-<text text-anchor="middle" x="294.6182" y="-800.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #deleter_</text>
+<path fill="none" stroke="#404040" d="M249.0442,-836.9248C259.7298,-830.6637 266.6816,-819.3555 266.6816,-803 266.6816,-792.0112 263.5435,-783.3007 258.2432,-776.8687"/>
+<polygon fill="none" stroke="#404040" points="258.2001,-776.8322 251.0365,-776.0056 249.0442,-769.0752 256.2078,-769.9017 258.2001,-776.8322"/>
+<text text-anchor="middle" x="292.6816" y="-800.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #deleter_</text>
</g>
<!-- Node7 -->
<g id="node6" class="node">
<title>Node7</title>
<g id="a_node6"><a xlink:href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html" target="_top" xlink:title="IRDocsifier is the top-level interface in the IR->Doc process. ">
-<polygon fill="#ffffff" stroke="#000000" points="352.6182,-193.5 352.6182,-382.5 561.6182,-382.5 561.6182,-193.5 352.6182,-193.5"/>
-<text text-anchor="start" x="360.6182" y="-370.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
-<text text-anchor="middle" x="457.1182" y="-359.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
-<polyline fill="none" stroke="#000000" points="352.6182,-352.5 561.6182,-352.5 "/>
-<text text-anchor="start" x="360.6182" y="-340.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ obj2info</text>
-<text text-anchor="start" x="360.6182" y="-329.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined_names</text>
-<text text-anchor="start" x="360.6182" y="-318.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ common_prefix</text>
-<text text-anchor="start" x="360.6182" y="-307.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<polyline fill="none" stroke="#000000" points="352.6182,-300.5 561.6182,-300.5 "/>
-<text text-anchor="start" x="360.6182" y="-288.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
-<text text-anchor="start" x="360.6182" y="-277.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
-<text text-anchor="start" x="360.6182" y="-266.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
-<text text-anchor="start" x="360.6182" y="-255.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
-<text text-anchor="start" x="360.6182" y="-244.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetVarDoc()</text>
-<text text-anchor="start" x="360.6182" y="-233.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsVarDefined()</text>
-<text text-anchor="start" x="360.6182" y="-222.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RemoveVar()</text>
-<text text-anchor="start" x="360.6182" y="-211.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ SetCommonPrefix()</text>
-<text text-anchor="start" x="360.6182" y="-200.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AsDoc()</text>
+<polygon fill="#ffffff" stroke="#000000" points="300.6816,-193.5 300.6816,-382.5 509.6816,-382.5 509.6816,-193.5 300.6816,-193.5"/>
+<text text-anchor="start" x="308.6816" y="-370.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
+<text text-anchor="middle" x="405.1816" y="-359.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
+<polyline fill="none" stroke="#000000" points="300.6816,-352.5 509.6816,-352.5 "/>
+<text text-anchor="start" x="308.6816" y="-340.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ obj2info</text>
+<text text-anchor="start" x="308.6816" y="-329.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined_names</text>
+<text text-anchor="start" x="308.6816" y="-318.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ common_prefix</text>
+<text text-anchor="start" x="308.6816" y="-307.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="300.6816,-300.5 509.6816,-300.5 "/>
+<text text-anchor="start" x="308.6816" y="-288.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
+<text text-anchor="start" x="308.6816" y="-277.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
+<text text-anchor="start" x="308.6816" y="-266.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
+<text text-anchor="start" x="308.6816" y="-255.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
+<text text-anchor="start" x="308.6816" y="-244.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetVarDoc()</text>
+<text text-anchor="start" x="308.6816" y="-233.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsVarDefined()</text>
+<text text-anchor="start" x="308.6816" y="-222.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RemoveVar()</text>
+<text text-anchor="start" x="308.6816" y="-211.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ SetCommonPrefix()</text>
+<text text-anchor="start" x="308.6816" y="-200.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AsDoc()</text>
</a>
</g>
</g>
<!-- Node3->Node7 -->
<g id="edge7" class="edge">
<title>Node3->Node7</title>
-<path fill="none" stroke="#191970" d="M255.8768,-635.7829C304.4169,-551.8966 361.4259,-453.3742 402.3499,-382.6499"/>
-<polygon fill="none" stroke="#191970" points="252.7354,-634.2237 250.7563,-644.632 258.7942,-637.7296 252.7354,-634.2237"/>
+<path fill="none" stroke="#191970" d="M253.2642,-603.4737C289.828,-527.5449 329.8075,-444.5229 359.5722,-382.7131"/>
+<polygon fill="none" stroke="#191970" points="250.0084,-602.1679 248.8231,-612.6962 256.3152,-605.205 250.0084,-602.1679"/>
</g>
<!-- Node4 -->
<g id="node3" class="node">
<title>Node4</title>
<g id="a_node3"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::script::printer\l::StmtDoc \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="2.6182,-401.5 2.6182,-590.5 121.6182,-590.5 121.6182,-401.5 2.6182,-401.5"/>
-<text text-anchor="start" x="10.6182" y="-578.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
-<text text-anchor="start" x="10.6182" y="-567.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
-<text text-anchor="middle" x="62.1182" y="-556.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::StmtDoc ></text>
-<polyline fill="none" stroke="#000000" points="2.6182,-549.5 121.6182,-549.5 "/>
-<text text-anchor="middle" x="62.1182" y="-537.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="2.6182,-530.5 121.6182,-530.5 "/>
-<text text-anchor="start" x="10.6182" y="-518.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-507.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-496.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-485.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-474.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-463.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-452.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-441.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="10.6182" y="-430.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="10.6182" y="-419.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="10.6182" y="-408.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
+<polygon fill="#ffffff" stroke="#000000" points=".6816,-401.5 .6816,-590.5 119.6816,-590.5 119.6816,-401.5 .6816,-401.5"/>
+<text text-anchor="start" x="8.6816" y="-578.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
+<text text-anchor="start" x="8.6816" y="-567.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
+<text text-anchor="middle" x="60.1816" y="-556.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::StmtDoc ></text>
+<polyline fill="none" stroke="#000000" points=".6816,-549.5 119.6816,-549.5 "/>
+<text text-anchor="middle" x="60.1816" y="-537.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points=".6816,-530.5 119.6816,-530.5 "/>
+<text text-anchor="start" x="8.6816" y="-518.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-507.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-496.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-485.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-474.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-463.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-452.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-441.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="8.6816" y="-430.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="8.6816" y="-419.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="8.6816" y="-408.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
</a>
</g>
</g>
<!-- Node4->Node2 -->
<g id="edge3" class="edge">
<title>Node4->Node2</title>
-<path fill="none" stroke="#404040" d="M68.6456,-401.3989C74.4994,-340.9001 85.3489,-261.2578 105.1182,-193 108.5618,-181.1101 112.9668,-168.8776 117.7719,-156.9801"/>
-<polygon fill="none" stroke="#404040" points="117.8591,-156.7714 116.48,-149.6934 122.4832,-145.6981 123.8622,-152.7761 117.8591,-156.7714"/>
-<text text-anchor="middle" x="125.1182" y="-285.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +stmts</text>
+<path fill="none" stroke="#404040" d="M66.7091,-401.3989C72.5629,-340.9001 83.4123,-261.2578 103.1816,-193 106.6253,-181.1101 111.0303,-168.8776 115.8354,-156.9801"/>
+<polygon fill="none" stroke="#404040" points="115.9225,-156.7714 114.5435,-149.6934 120.5467,-145.6981 121.9257,-152.7761 115.9225,-156.7714"/>
+<text text-anchor="middle" x="123.1816" y="-285.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +stmts</text>
</g>
<!-- Node5 -->
<g id="node4" class="node">
<title>Node5</title>
<g id="a_node4"><a xlink:href="classtvm_1_1runtime_1_1ObjectRef.html" target="_top" xlink:title="Base class of all object reference. ">
-<polygon fill="#ffffff" stroke="#000000" points="487.1182,-1034.5 487.1182,-1256.5 621.1182,-1256.5 621.1182,-1034.5 487.1182,-1034.5"/>
-<text text-anchor="middle" x="554.1182" y="-1244.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectRef</text>
-<polyline fill="none" stroke="#000000" points="487.1182,-1237.5 621.1182,-1237.5 "/>
-<text text-anchor="start" x="495.1182" y="-1225.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
-<polyline fill="none" stroke="#000000" points="487.1182,-1218.5 621.1182,-1218.5 "/>
-<text text-anchor="start" x="495.1182" y="-1206.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
-<text text-anchor="start" x="495.1182" y="-1195.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
-<text text-anchor="start" x="495.1182" y="-1184.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ same_as()</text>
-<text text-anchor="start" x="495.1182" y="-1173.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator==()</text>
-<text text-anchor="start" x="495.1182" y="-1162.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator!=()</text>
-<text text-anchor="start" x="495.1182" y="-1151.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator<()</text>
-<text text-anchor="start" x="495.1182" y="-1140.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined()</text>
-<text text-anchor="start" x="495.1182" y="-1129.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
-<text text-anchor="start" x="495.1182" y="-1118.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
-<text text-anchor="start" x="495.1182" y="-1107.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
-<text text-anchor="start" x="495.1182" y="-1096.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ use_count()</text>
-<text text-anchor="start" x="495.1182" y="-1085.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ as()</text>
-<text text-anchor="start" x="495.1182" y="-1074.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># get_mutable()</text>
-<text text-anchor="start" x="495.1182" y="-1063.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DowncastNoCheck()</text>
-<text text-anchor="start" x="495.1182" y="-1052.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># FFIClearAfterMove()</text>
-<text text-anchor="start" x="495.1182" y="-1041.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetDataPtr()</text>
+<polygon fill="#ffffff" stroke="#000000" points="410.1816,-1034.5 410.1816,-1256.5 544.1816,-1256.5 544.1816,-1034.5 410.1816,-1034.5"/>
+<text text-anchor="middle" x="477.1816" y="-1244.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectRef</text>
+<polyline fill="none" stroke="#000000" points="410.1816,-1237.5 544.1816,-1237.5 "/>
+<text text-anchor="start" x="418.1816" y="-1225.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
+<polyline fill="none" stroke="#000000" points="410.1816,-1218.5 544.1816,-1218.5 "/>
+<text text-anchor="start" x="418.1816" y="-1206.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
+<text text-anchor="start" x="418.1816" y="-1195.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
+<text text-anchor="start" x="418.1816" y="-1184.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ same_as()</text>
+<text text-anchor="start" x="418.1816" y="-1173.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator==()</text>
+<text text-anchor="start" x="418.1816" y="-1162.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator!=()</text>
+<text text-anchor="start" x="418.1816" y="-1151.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator<()</text>
+<text text-anchor="start" x="418.1816" y="-1140.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined()</text>
+<text text-anchor="start" x="418.1816" y="-1129.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
+<text text-anchor="start" x="418.1816" y="-1118.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
+<text text-anchor="start" x="418.1816" y="-1107.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
+<text text-anchor="start" x="418.1816" y="-1096.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ use_count()</text>
+<text text-anchor="start" x="418.1816" y="-1085.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ as()</text>
+<text text-anchor="start" x="418.1816" y="-1074.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># get_mutable()</text>
+<text text-anchor="start" x="418.1816" y="-1063.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DowncastNoCheck()</text>
+<text text-anchor="start" x="418.1816" y="-1052.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># FFIClearAfterMove()</text>
+<text text-anchor="start" x="418.1816" y="-1041.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetDataPtr()</text>
</a>
</g>
</g>
<!-- Node5->Node4 -->
<g id="edge4" class="edge">
<title>Node5->Node4</title>
-<path fill="none" stroke="#191970" d="M476.7483,-1133.7047C350.298,-1112.6713 108.3545,-1064.48 58.1182,-997 -29.2734,-879.611 -.0651,-700.119 30.2266,-590.8613"/>
-<polygon fill="none" stroke="#191970" points="476.3255,-1137.1821 486.7614,-1135.3536 477.463,-1130.2752 476.3255,-1137.1821"/>
+<path fill="none" stroke="#191970" d="M400.1166,-1129.4137C290.0647,-1104.7006 96.6659,-1054.209 56.1816,-997 -28.2299,-877.7165 -.1287,-699.2299 29.1654,-590.6154"/>
+<polygon fill="none" stroke="#191970" points="399.6511,-1132.8956 410.1727,-1131.6522 401.1722,-1126.0629 399.6511,-1132.8956"/>
</g>
<!-- Node8 -->
<g id="node7" class="node">
<title>Node8</title>
<g id="a_node7"><a xlink:href="classtvm_1_1runtime_1_1Optional.html" target="_top" xlink:title="{tvm::runtime::Optional\l\< tvm::IRModule \>\n|+ _type_is_nullable\l|+ Optional()\l+ Optional()\l+ Optional()\l+ Optional()\l+ Optional()\l+ Optional()\l+ Optional()\l+ Optional()\l+ operator=()\l+ operator=()\land 15 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="778.1182,-714 778.1182,-892 906.1182,-892 906.1182,-714 778.1182,-714"/>
-<text text-anchor="start" x="786.1182" y="-880" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Optional</text>
-<text text-anchor="middle" x="842.1182" y="-869" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::IRModule ></text>
-<polyline fill="none" stroke="#000000" points="778.1182,-862 906.1182,-862 "/>
-<text text-anchor="start" x="786.1182" y="-850" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
-<polyline fill="none" stroke="#000000" points="778.1182,-843 906.1182,-843 "/>
-<text text-anchor="start" x="786.1182" y="-831" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-820" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-809" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-798" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-787" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-776" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-765" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-754" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
-<text text-anchor="start" x="786.1182" y="-743" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="786.1182" y="-732" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="786.1182" y="-721" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 15 more...</text>
+<polygon fill="#ffffff" stroke="#000000" points="629.1816,-714 629.1816,-892 757.1816,-892 757.1816,-714 629.1816,-714"/>
+<text text-anchor="start" x="637.1816" y="-880" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Optional</text>
+<text text-anchor="middle" x="693.1816" y="-869" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::IRModule ></text>
+<polyline fill="none" stroke="#000000" points="629.1816,-862 757.1816,-862 "/>
+<text text-anchor="start" x="637.1816" y="-850" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
+<polyline fill="none" stroke="#000000" points="629.1816,-843 757.1816,-843 "/>
+<text text-anchor="start" x="637.1816" y="-831" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-820" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-809" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-798" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-787" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-776" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-765" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-754" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Optional()</text>
+<text text-anchor="start" x="637.1816" y="-743" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="637.1816" y="-732" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="637.1816" y="-721" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 15 more...</text>
</a>
</g>
</g>
<!-- Node5->Node8 -->
<g id="edge9" class="edge">
<title>Node5->Node8</title>
-<path fill="none" stroke="#191970" d="M630.1985,-1108.1914C675.8885,-1082.4405 732.1389,-1044.3416 769.1182,-997 792.9319,-966.5133 809.6182,-927.4217 820.9639,-892.328"/>
-<polygon fill="none" stroke="#191970" points="628.258,-1105.2653 621.2102,-1113.176 631.6529,-1111.387 628.258,-1105.2653"/>
+<path fill="none" stroke="#191970" d="M551.7643,-1078.9638C575.6122,-1054.8979 600.6128,-1026.3746 619.1816,-997 639.4061,-965.0063 655.4891,-926.6212 667.4255,-892.4699"/>
+<polygon fill="none" stroke="#191970" points="549.027,-1076.7502 544.4086,-1086.2855 553.9652,-1081.7114 549.027,-1076.7502"/>
</g>
<!-- Node9 -->
<g id="node8" class="node">
<title>Node9</title>
-<g id="a_node8"><a xlink:href="classtvm_1_1runtime_1_1Map.html" target="_top" xlink:title="{tvm::runtime::Map\<\l tvm::runtime::String,\l tvm::runtime::String \>\n||+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ operator=()\l+ operator=()\l+ at()\land 12 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="338.6182,-708.5 338.6182,-897.5 467.6182,-897.5 467.6182,-708.5 338.6182,-708.5"/>
-<text text-anchor="start" x="346.6182" y="-885.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Map<</text>
-<text text-anchor="start" x="346.6182" y="-874.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> tvm::runtime::String,</text>
-<text text-anchor="middle" x="403.1182" y="-863.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> tvm::runtime::String ></text>
-<polyline fill="none" stroke="#000000" points="338.6182,-856.5 467.6182,-856.5 "/>
-<text text-anchor="middle" x="403.1182" y="-844.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="338.6182,-837.5 467.6182,-837.5 "/>
-<text text-anchor="start" x="346.6182" y="-825.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-814.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-803.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-792.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-781.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-770.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-759.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="346.6182" y="-748.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="346.6182" y="-737.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="346.6182" y="-726.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ at()</text>
-<text text-anchor="start" x="346.6182" y="-715.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 12 more...</text>
+<g id="a_node8"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::runtime::String \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
+<polygon fill="#ffffff" stroke="#000000" points="336.6816,-714 336.6816,-892 473.6816,-892 473.6816,-714 336.6816,-714"/>
+<text text-anchor="start" x="344.6816" y="-880" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
+<text text-anchor="middle" x="405.1816" y="-869" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::String ></text>
+<polyline fill="none" stroke="#000000" points="336.6816,-862 473.6816,-862 "/>
+<text text-anchor="middle" x="405.1816" y="-850" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="336.6816,-843 473.6816,-843 "/>
+<text text-anchor="start" x="344.6816" y="-831" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-820" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-809" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-798" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-787" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-776" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-765" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-754" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="344.6816" y="-743" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="344.6816" y="-732" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="344.6816" y="-721" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
</a>
</g>
</g>
<!-- Node5->Node9 -->
<g id="edge11" class="edge">
<title>Node5->Node9</title>
-<path fill="none" stroke="#191970" d="M489.3664,-1025.2443C484.7443,-1015.7549 480.274,-1006.2619 476.1182,-997 461.8282,-965.1525 447.8868,-929.5869 436.0813,-897.6931"/>
-<polygon fill="none" stroke="#191970" points="486.2835,-1026.9065 493.8391,-1034.3337 492.5642,-1023.8159 486.2835,-1026.9065"/>
+<path fill="none" stroke="#191970" d="M451.7461,-1024.5043C442.5894,-980.9464 432.4322,-932.6293 423.947,-892.2656"/>
+<polygon fill="none" stroke="#191970" points="448.3229,-1025.234 453.8053,-1034.3001 455.1731,-1023.7939 448.3229,-1025.234"/>
</g>
<!-- Node10 -->
<g id="node9" class="node">
<title>Node10</title>
-<g id="a_node9"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::runtime::String \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="485.6182,-714 485.6182,-892 622.6182,-892 622.6182,-714 485.6182,-714"/>
-<text text-anchor="start" x="493.6182" y="-880" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
-<text text-anchor="middle" x="554.1182" y="-869" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::String ></text>
-<polyline fill="none" stroke="#000000" points="485.6182,-862 622.6182,-862 "/>
-<text text-anchor="middle" x="554.1182" y="-850" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="485.6182,-843 622.6182,-843 "/>
-<text text-anchor="start" x="493.6182" y="-831" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-820" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-809" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-798" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-787" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-776" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-765" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-754" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="493.6182" y="-743" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="493.6182" y="-732" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="493.6182" y="-721" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
+<g id="a_node9"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::script::printer\l::Frame \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
+<polygon fill="#ffffff" stroke="#000000" points="491.6816,-708.5 491.6816,-897.5 610.6816,-897.5 610.6816,-708.5 491.6816,-708.5"/>
+<text text-anchor="start" x="499.6816" y="-885.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
+<text text-anchor="start" x="499.6816" y="-874.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
+<text text-anchor="middle" x="551.1816" y="-863.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::Frame ></text>
+<polyline fill="none" stroke="#000000" points="491.6816,-856.5 610.6816,-856.5 "/>
+<text text-anchor="middle" x="551.1816" y="-844.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="491.6816,-837.5 610.6816,-837.5 "/>
+<text text-anchor="start" x="499.6816" y="-825.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-814.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-803.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-792.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-781.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-770.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-759.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-748.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="499.6816" y="-737.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="499.6816" y="-726.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="499.6816" y="-715.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
</a>
</g>
</g>
<!-- Node5->Node10 -->
<g id="edge13" class="edge">
<title>Node5->Node10</title>
-<path fill="none" stroke="#191970" d="M554.1182,-1024.2298C554.1182,-980.7434 554.1182,-932.5445 554.1182,-892.2656"/>
-<polygon fill="none" stroke="#191970" points="550.6183,-1024.3 554.1182,-1034.3001 557.6183,-1024.3001 550.6183,-1024.3"/>
-</g>
-<!-- Node11 -->
-<g id="node10" class="node">
-<title>Node11</title>
-<g id="a_node10"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::script::printer\l::Frame \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="640.6182,-708.5 640.6182,-897.5 759.6182,-897.5 759.6182,-708.5 640.6182,-708.5"/>
-<text text-anchor="start" x="648.6182" y="-885.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
-<text text-anchor="start" x="648.6182" y="-874.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
-<text text-anchor="middle" x="700.1182" y="-863.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::Frame ></text>
-<polyline fill="none" stroke="#000000" points="640.6182,-856.5 759.6182,-856.5 "/>
-<text text-anchor="middle" x="700.1182" y="-844.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="640.6182,-837.5 759.6182,-837.5 "/>
-<text text-anchor="start" x="648.6182" y="-825.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-814.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-803.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-792.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-781.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-770.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-759.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-748.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="648.6182" y="-737.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="648.6182" y="-726.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="648.6182" y="-715.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
-</a>
-</g>
-</g>
-<!-- Node5->Node11 -->
-<g id="edge15" class="edge">
-<title>Node5->Node11</title>
-<path fill="none" stroke="#191970" d="M618.1896,-1025.2946C622.7182,-1015.7937 627.0825,-1006.2843 631.1182,-997 644.9411,-965.1993 658.1951,-929.645 669.3269,-897.7472"/>
-<polygon fill="none" stroke="#191970" points="614.9929,-1023.8658 613.8027,-1034.3936 621.2983,-1026.9059 614.9929,-1023.8658"/>
+<path fill="none" stroke="#191970" d="M503.3515,-1024.3761C512.3469,-982.742 522.2788,-936.7731 530.7504,-897.5637"/>
+<polygon fill="none" stroke="#191970" points="499.8982,-1023.7864 501.2073,-1034.3001 506.7403,-1025.2648 499.8982,-1023.7864"/>
</g>
<!-- Node6 -->
<g id="node5" class="node">
<title>Node6</title>
<g id="a_node5"><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="484.1182,-1304.5 484.1182,-1482.5 624.1182,-1482.5 624.1182,-1304.5 484.1182,-1304.5"/>
-<text text-anchor="start" x="492.1182" y="-1470.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectPtr</text>
-<text text-anchor="middle" x="554.1182" y="-1459.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::Object ></text>
-<polyline fill="none" stroke="#000000" points="484.1182,-1452.5 624.1182,-1452.5 "/>
-<text text-anchor="middle" x="554.1182" y="-1440.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="484.1182,-1433.5 624.1182,-1433.5 "/>
-<text text-anchor="start" x="492.1182" y="-1421.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1410.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1399.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1388.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1377.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1366.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1355.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~ObjectPtr()</text>
-<text text-anchor="start" x="492.1182" y="-1344.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ swap()</text>
-<text text-anchor="start" x="492.1182" y="-1333.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
-<text text-anchor="start" x="492.1182" y="-1322.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
-<text text-anchor="start" x="492.1182" y="-1311.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 11 more...</text>
+<polygon fill="#ffffff" stroke="#000000" points="407.1816,-1304.5 407.1816,-1482.5 547.1816,-1482.5 547.1816,-1304.5 407.1816,-1304.5"/>
+<text text-anchor="start" x="415.1816" y="-1470.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectPtr</text>
+<text text-anchor="middle" x="477.1816" y="-1459.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::Object ></text>
+<polyline fill="none" stroke="#000000" points="407.1816,-1452.5 547.1816,-1452.5 "/>
+<text text-anchor="middle" x="477.1816" y="-1440.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="407.1816,-1433.5 547.1816,-1433.5 "/>
+<text text-anchor="start" x="415.1816" y="-1421.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1410.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1399.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1388.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1377.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1366.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1355.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~ObjectPtr()</text>
+<text text-anchor="start" x="415.1816" y="-1344.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ swap()</text>
+<text text-anchor="start" x="415.1816" y="-1333.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
+<text text-anchor="start" x="415.1816" y="-1322.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
+<text text-anchor="start" x="415.1816" y="-1311.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 11 more...</text>
</a>
</g>
</g>
<!-- Node6->Node5 -->
<g id="edge5" class="edge">
<title>Node6->Node5</title>
-<path fill="none" stroke="#404040" d="M554.1182,-1304.3167C554.1182,-1292.8765 554.1182,-1281.0062 554.1182,-1269.1402"/>
-<polygon fill="none" stroke="#404040" points="554.1183,-1268.7944 550.1182,-1262.7944 554.1182,-1256.7944 558.1182,-1262.7943 554.1183,-1268.7944"/>
-<text text-anchor="middle" x="573.6182" y="-1278" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #data_</text>
+<path fill="none" stroke="#404040" d="M477.1816,-1304.3167C477.1816,-1292.8765 477.1816,-1281.0062 477.1816,-1269.1402"/>
+<polygon fill="none" stroke="#404040" points="477.1817,-1268.7944 473.1817,-1262.7944 477.1816,-1256.7944 481.1817,-1262.7943 477.1817,-1268.7944"/>
+<text text-anchor="middle" x="496.6816" y="-1278" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #data_</text>
</g>
<!-- Node7->Node2 -->
<g id="edge6" class="edge">
<title>Node7->Node2</title>
-<path fill="none" stroke="#404040" d="M352.4736,-212.5014C325.7375,-193.2119 296.9049,-172.4099 269.8736,-152.9074"/>
-<polygon fill="none" stroke="#404040" points="269.735,-152.8074 262.5288,-152.5407 260.0033,-145.7863 267.2095,-146.053 269.735,-152.8074"/>
-<text text-anchor="middle" x="309.1182" y="-167" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +d</text>
+<path fill="none" stroke="#404040" d="M300.3668,-197.1323C283.7905,-182.7617 266.7455,-167.9848 250.4556,-153.8625"/>
+<polygon fill="none" stroke="#404040" points="250.0279,-153.4917 242.8741,-152.5838 240.9608,-145.6311 248.1145,-146.539 250.0279,-153.4917"/>
+<text text-anchor="middle" x="283.1816" y="-167" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +d</text>
</g>
<!-- Node8->Node7 -->
<g id="edge8" class="edge">
<title>Node8->Node7</title>
-<path fill="none" stroke="#404040" d="M843.3162,-713.5929C839.7887,-624.252 820.9423,-488.0066 745.1182,-401 701.532,-350.9857 632.9615,-322.9283 573.6622,-307.2829"/>
-<polygon fill="none" stroke="#404040" points="573.5696,-307.2595 566.7718,-309.6658 561.9361,-304.3164 568.7339,-301.9101 573.5696,-307.2595"/>
-<text text-anchor="middle" x="844.1182" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +mod</text>
+<path fill="none" stroke="#404040" d="M691.2445,-713.9504C685.1614,-626.4692 664.2611,-493.0776 596.1816,-401 576.5973,-374.5121 548.7894,-353.1126 520.4357,-336.3573"/>
+<polygon fill="none" stroke="#404040" points="520.403,-336.3386 513.2093,-336.8407 509.9791,-330.3935 517.1727,-329.8915 520.403,-336.3386"/>
+<text text-anchor="middle" x="690.1816" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +mod</text>
</g>
<!-- Node9->Node7 -->
<g id="edge10" class="edge">
<title>Node9->Node7</title>
-<path fill="none" stroke="#404040" d="M413.0541,-708.2411C422.3537,-619.5503 436.2298,-487.2134 445.9183,-394.8142"/>
-<polygon fill="none" stroke="#404040" points="445.9456,-394.5522 442.5932,-388.1678 447.1971,-382.6176 450.5496,-389.0021 445.9456,-394.5522"/>
-<text text-anchor="middle" x="470.1182" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +ir_prefix</text>
+<path fill="none" stroke="#404040" d="M405.1816,-713.8818C405.1816,-625.2118 405.1816,-489.1062 405.1816,-394.7234"/>
+<polygon fill="none" stroke="#404040" points="405.1817,-394.5602 401.1817,-388.5603 405.1816,-382.5602 409.1817,-388.5602 405.1817,-394.5602"/>
+<text text-anchor="middle" x="450.6816" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +dispatch_tokens</text>
</g>
<!-- Node10->Node7 -->
<g id="edge12" class="edge">
<title>Node10->Node7</title>
-<path fill="none" stroke="#404040" d="M548.0399,-713.6584C541.0569,-631.4449 526.9274,-506.6234 500.1182,-401 499.5707,-398.8431 498.9949,-396.671 498.3939,-394.4888"/>
-<polygon fill="none" stroke="#404040" points="498.3289,-394.2665 492.805,-389.6311 494.9592,-382.7493 500.4831,-387.3847 498.3289,-394.2665"/>
-<text text-anchor="middle" x="580.6182" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +dispatch_tokens</text>
-</g>
-<!-- Node11->Node7 -->
-<g id="edge14" class="edge">
-<title>Node11->Node7</title>
-<path fill="none" stroke="#404040" d="M704.2338,-708.3176C703.5111,-621.5609 690.6993,-493.0051 630.1182,-401 615.2527,-378.4236 594.1506,-359.3478 571.771,-343.6676"/>
-<polygon fill="none" stroke="#404040" points="571.7204,-343.6335 564.5093,-343.5989 561.7678,-336.9292 568.9789,-336.9639 571.7204,-343.6335"/>
-<text text-anchor="middle" x="718.1182" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +frames</text>
+<path fill="none" stroke="#404040" d="M554.0243,-708.4355C553.1647,-624.3425 543.4171,-499.7641 500.1816,-401 499.0326,-398.3752 497.8027,-395.7641 496.5013,-393.1705"/>
+<polygon fill="none" stroke="#404040" points="496.4972,-393.1629 490.119,-389.7985 490.7715,-382.6169 497.1496,-385.9814 496.4972,-393.1629"/>
+<text text-anchor="middle" x="569.1816" y="-493.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +frames</text>
</g>
</g>
</svg>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier-members.html b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier-members.html
index 17e0a43c98..afeaba21ea 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier-members.html
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier-members.html
@@ -80,7 +80,7 @@ $(function() {
<tr class="even"><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><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 class="even"><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><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a384e4e9461de539e28bb27efeefc9e19">IRDocsifier</a>(Map< String, String > ir_prefix)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html">tvm::script::printer::IRDocsifier</a></td><td class="entry"><span class="mlabel">explicit</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a495e99eb65db294cbc6afb33119266fd">IRDocsifier</a>()</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html">tvm::script::printer::IRDocsifier</a></td><td class="entry"></td></tr>
<tr class="even"><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><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 class="even"><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>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier.html b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier.html
index 6b98e43f98..7470d27fa8 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier.html
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifier.html
@@ -100,9 +100,9 @@ Public Types</h2></td></tr>
</table><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:a384e4e9461de539e28bb27efeefc9e19"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a384e4e9461de539e28bb27efeefc9e19">IRDocsifier</a> (<a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a>< <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a>, <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a> > ir_prefi [...]
-<tr class="memdesc:a384e4e9461de539e28bb27efeefc9e19"><td class="mdescLeft"> </td><td class="mdescRight">Create a <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" title="Reference type of IRDocsifierNode. ">IRDocsifier</a>. <a href="#a384e4e9461de539e28bb27efeefc9e19">More...</a><br /></td></tr>
-<tr class="separator:a384e4e9461de539e28bb27efeefc9e19"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:a495e99eb65db294cbc6afb33119266fd"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a495e99eb65db294cbc6afb33119266fd">IRDocsifier</a> ()</td></tr>
+<tr class="memdesc:a495e99eb65db294cbc6afb33119266fd"><td class="mdescLeft"> </td><td class="mdescRight">Create a <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" title="Reference type of IRDocsifierNode. ">IRDocsifier</a>. <a href="#a495e99eb65db294cbc6afb33119266fd">More...</a><br /></td></tr>
+<tr class="separator:a495e99eb65db294cbc6afb33119266fd"><td class="memSeparator" colspan="2"> </td></tr>
<tr class="memitem:abd765959d8e2f7e20969f29e8b803c31"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#abd765959d8e2f7e20969f29e8b803c31">TVM_DEFINE_MUTABLE_NOTNULLABLE_OBJECT_REF_METHODS</a> (<a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html">IRDocsifier</a>, <a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html">ObjectRef</a>, <a class="el" [...]
<tr class="separator:abd765959d8e2f7e20969f29e8b803c31"><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>
@@ -188,37 +188,22 @@ Additional Inherited Members</h2></td></tr>
</div>
</div>
<h2 class="groupheader">Constructor & Destructor Documentation</h2>
-<a id="a384e4e9461de539e28bb27efeefc9e19"></a>
-<h2 class="memtitle"><span class="permalink"><a href="#a384e4e9461de539e28bb27efeefc9e19">◆ </a></span>IRDocsifier()</h2>
+<a id="a495e99eb65db294cbc6afb33119266fd"></a>
+<h2 class="memtitle"><span class="permalink"><a href="#a495e99eb65db294cbc6afb33119266fd">◆ </a></span>IRDocsifier()</h2>
<div class="memitem">
<div class="memproto">
-<table class="mlabels">
- <tr>
- <td class="mlabels-left">
<table class="memname">
<tr>
<td class="memname">tvm::script::printer::IRDocsifier::IRDocsifier </td>
<td>(</td>
- <td class="paramtype"><a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a>< <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a>, <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a> > </td>
- <td class="paramname"><em>ir_prefix</em></td><td>)</td>
+ <td class="paramname"></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">
<p>Create a <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" title="Reference type of IRDocsifierNode. ">IRDocsifier</a>. </p>
-<dl class="params"><dt>Parameters</dt><dd>
- <table class="params">
- <tr><td class="paramname">ir_prefix</td><td>The ir_prefix to use for this <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" title="Reference type of IRDocsifierNode. ">IRDocsifier</a>. </td></tr>
- </table>
- </dd>
-</dl>
</div>
</div>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode-members.html b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode-members.html
index 25d5207b9c..04e8da5d44 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode-members.html
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode-members.html
@@ -94,29 +94,28 @@ $(function() {
<tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a5693cbadcc1168b96db7b1cc5c200b86">GetTypeKeyHash</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
<tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a30f58189ba1dd28c8a99c154d505c123">GetVarDoc</a>(const ObjectRef &obj) const</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
<tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#ac9e5eed7719e322117bde996a171e33a">IncRef</a>()</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1script_1_1printer_1_1IRDocsifierNode.html#adccf933a99d410b630f223ac1afe548f">ir_prefix</a></td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a90e90b3f4ba8a590baff78c75807bbc7">IsInstance</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a972f145d5a6c7e8f63f009e63ea76af0">IsVarDefined</a>(const ObjectRef &obj) const</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#af5cf9a40e41ba47125ad2b69deeab92a">mod</a></td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a5ce74fda8a1c8dbcdf47537540c2219f">obj2info</a></td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a133436a9ec5c4a768b94102bf95a660b">Object</a>()</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#ab7968feb6ad38ecaffc320e13819d826">Object</a>(const Object &other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#aa1612f69ea5b4225d4cda759cd517323">Object</a>(Object &&other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a69c32fbd96181f5c21d2c878ab285e4f">operator=</a>(const Object &other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#ae341e561272ff43cdcbc927bc29ac50d">operator=</a>(Object &&other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a0d492efee331e2239a093f4b2017c10f">ref_counter_</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#a55549a6c23987890246248682560a03d">RefCounterType</a> typedef</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#ad6bbe086197861a45275303dd5f0db99">RemoveVar</a>(const ObjectRef &obj)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#ad94d79729ac85aa7c976e23d39066383">RuntimeTypeIndex</a>()</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">static</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#adb1e2b064e6c76df9fa2cc8656a73311">SetCommonPrefix</a>(const ObjectRef &root, runtime::TypedPackedFunc< bool(ObjectRef)> is_var)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a0be0d429c6a05835b380b23905dacaf7">TVM_DECLARE_FINAL_OBJECT_INFO</a>(IRDocsifierNode, Object)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a481f01923b14e1851ebd38506e9c66ea">type_index</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#a4bfc2586cb55f2af47728187b3256255">type_index_</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">protected</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a817ba6c23b7ee1821c48a75edf255a30">TypeIndex2Key</a>(uint32_t tindex)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a6ee32a02dd44257da105fbbe5d9c8622">TypeIndex2KeyHash</a>(uint32_t tindex)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a6841f97e06e6614dd7e82c6dd41b818a">TypeKey2Index</a>(const std::string &key)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
- <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#afd548730a6139d19fe24473ad66026d7">unique</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
- <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#ae651362767ff31ca0ccc8f37b0fe6dc0">VisitAttrs</a>(tvm::AttrVisitor *v)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a90e90b3f4ba8a590baff78c75807bbc7">IsInstance</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a972f145d5a6c7e8f63f009e63ea76af0">IsVarDefined</a>(const ObjectRef &obj) const</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#af5cf9a40e41ba47125ad2b69deeab92a">mod</a></td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a5ce74fda8a1c8dbcdf47537540c2219f">obj2info</a></td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a133436a9ec5c4a768b94102bf95a660b">Object</a>()</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#ab7968feb6ad38ecaffc320e13819d826">Object</a>(const Object &other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#aa1612f69ea5b4225d4cda759cd517323">Object</a>(Object &&other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#a69c32fbd96181f5c21d2c878ab285e4f">operator=</a>(const Object &other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#ae341e561272ff43cdcbc927bc29ac50d">operator=</a>(Object &&other)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#a0d492efee331e2239a093f4b2017c10f">ref_counter_</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">protected</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a55549a6c23987890246248682560a03d">RefCounterType</a> typedef</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#ad6bbe086197861a45275303dd5f0db99">RemoveVar</a>(const ObjectRef &obj)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#ad94d79729ac85aa7c976e23d39066383">RuntimeTypeIndex</a>()</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span><span class="mlabel">static</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#adb1e2b064e6c76df9fa2cc8656a73311">SetCommonPrefix</a>(const ObjectRef &root, runtime::TypedPackedFunc< bool(ObjectRef)> is_var)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a0be0d429c6a05835b380b23905dacaf7">TVM_DECLARE_FINAL_OBJECT_INFO</a>(IRDocsifierNode, Object)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</a></td><td class="entry"></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a481f01923b14e1851ebd38506e9c66ea">type_index</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a4bfc2586cb55f2af47728187b3256255">type_index_</a></td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</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_1Object.html#a817ba6c23b7ee1821c48a75edf255a30">TypeIndex2Key</a>(uint32_t tindex)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a6ee32a02dd44257da105fbbe5d9c8622">TypeIndex2KeyHash</a>(uint32_t tindex)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#a6841f97e06e6614dd7e82c6dd41b818a">TypeKey2Index</a>(const std::string &key)</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">static</span></td></tr>
+ <tr><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html#afd548730a6139d19fe24473ad66026d7">unique</a>() const</td><td class="entry"><a class="el" href="classtvm_1_1runtime_1_1Object.html">tvm::runtime::Object</a></td><td class="entry"><span class="mlabel">inline</span></td></tr>
+ <tr class="even"><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#ae651362767ff31ca0ccc8f37b0fe6dc0">VisitAttrs</a>(tvm::AttrVisitor *v)</td><td class="entry"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html">tvm::script::printer::IRDocsifierNode</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>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html
index 45c5324b95..b3945aa793 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html
@@ -81,13 +81,13 @@ $(function() {
<div class="dynheader">
Inheritance diagram for tvm::script::printer::IRDocsifierNode:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg" width="290" height="903"><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_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg" width="290" height="888"><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::script::printer::IRDocsifierNode:</div>
<div class="dyncontent">
-<div class="center"><iframe scrolling="no" frameborder="0" src="classtvm_1_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg" width="1130" height="1492"><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_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg" width="934" height="1492"><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">
@@ -162,9 +162,6 @@ Public Member Functions</h2></td></tr>
</table><table class="memberdecls">
<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-attribs"></a>
Public Attributes</h2></td></tr>
-<tr class="memitem:adccf933a99d410b630f223ac1afe548f"><td class="memItemLeft" align="right" valign="top"><a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a>< <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a>, <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a> > </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#adccf933a99d410b630f223ac1afe548f">ir_prefix</a></td></tr>
-<tr class="memdesc:adccf933a99d410b630f223ac1afe548f"><td class="mdescLeft"> </td><td class="mdescRight">This map connects IR dispatch token to the name of identifier. <a href="#adccf933a99d410b630f223ac1afe548f">More...</a><br /></td></tr>
-<tr class="separator:adccf933a99d410b630f223ac1afe548f"><td class="memSeparator" colspan="2"> </td></tr>
<tr class="memitem:a9a7d03d50540d3dd488a80975839f17d"><td class="memItemLeft" align="right" valign="top"><a class="el" href="classtvm_1_1runtime_1_1Array.html">Array</a>< <a class="el" href="classtvm_1_1script_1_1printer_1_1Frame.html">Frame</a> > </td><td class="memItemRight" valign="bottom"><a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#a9a7d03d50540d3dd488a80975839f17d">frames</a></td></tr>
<tr class="memdesc:a9a7d03d50540d3dd488a80975839f17d"><td class="mdescLeft"> </td><td class="mdescRight">The stack of frames. <a href="#a9a7d03d50540d3dd488a80975839f17d">More...</a><br /></td></tr>
<tr class="separator:a9a7d03d50540d3dd488a80975839f17d"><td class="memSeparator" colspan="2"> </td></tr>
@@ -657,22 +654,6 @@ template<class TDoc > </div>
<p>The stack of frames. </p>
<dl class="section see"><dt>See also</dt><dd><a class="el" href="classtvm_1_1script_1_1printer_1_1FrameNode.html">FrameNode</a> </dd></dl>
-</div>
-</div>
-<a id="adccf933a99d410b630f223ac1afe548f"></a>
-<h2 class="memtitle"><span class="permalink"><a href="#adccf933a99d410b630f223ac1afe548f">◆ </a></span>ir_prefix</h2>
-
-<div class="memitem">
-<div class="memproto">
- <table class="memname">
- <tr>
- <td class="memname"><a class="el" href="classtvm_1_1runtime_1_1Map.html">Map</a><<a class="el" href="classtvm_1_1runtime_1_1String.html">String</a>, <a class="el" href="classtvm_1_1runtime_1_1String.html">String</a>> tvm::script::printer::IRDocsifierNode::ir_prefix</td>
- </tr>
- </table>
-</div><div class="memdoc">
-
-<p>This map connects IR dispatch token to the name of identifier. </p>
-
</div>
</div>
<a id="af5cf9a40e41ba47125ad2b69deeab92a"></a>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg
index 09444b2990..3a1a0a7bb7 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__coll__graph.svg
@@ -4,32 +4,32 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: tvm::script::printer::IRDocsifierNode Pages: 1 -->
-<svg width="847pt" height="1119pt"
- viewBox="0.00 0.00 847.00 1119.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<svg width="700pt" height="1119pt"
+ viewBox="0.00 0.00 700.00 1119.00" 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 1115)">
<title>tvm::script::printer::IRDocsifierNode</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-1115 843,-1115 843,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-1115 696,-1115 696,4 -4,4"/>
<!-- Node2 -->
<g id="node1" class="node">
<title>Node2</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="378,-.5 378,-189.5 587,-189.5 587,-.5 378,-.5"/>
-<text text-anchor="start" x="386" y="-177.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
-<text text-anchor="middle" x="482.5" y="-166.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
-<polyline fill="none" stroke="#000000" points="378,-159.5 587,-159.5 "/>
-<text text-anchor="start" x="386" y="-147.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ obj2info</text>
-<text text-anchor="start" x="386" y="-136.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined_names</text>
-<text text-anchor="start" x="386" y="-125.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ common_prefix</text>
-<text text-anchor="start" x="386" y="-114.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<polyline fill="none" stroke="#000000" points="378,-107.5 587,-107.5 "/>
-<text text-anchor="start" x="386" y="-95.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
-<text text-anchor="start" x="386" y="-84.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
-<text text-anchor="start" x="386" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
-<text text-anchor="start" x="386" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
-<text text-anchor="start" x="386" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetVarDoc()</text>
-<text text-anchor="start" x="386" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsVarDefined()</text>
-<text text-anchor="start" x="386" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RemoveVar()</text>
-<text text-anchor="start" x="386" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ SetCommonPrefix()</text>
-<text text-anchor="start" x="386" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AsDoc()</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="303,-.5 303,-189.5 512,-189.5 512,-.5 303,-.5"/>
+<text text-anchor="start" x="311" y="-177.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
+<text text-anchor="middle" x="407.5" y="-166.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
+<polyline fill="none" stroke="#000000" points="303,-159.5 512,-159.5 "/>
+<text text-anchor="start" x="311" y="-147.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ obj2info</text>
+<text text-anchor="start" x="311" y="-136.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined_names</text>
+<text text-anchor="start" x="311" y="-125.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ common_prefix</text>
+<text text-anchor="start" x="311" y="-114.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<polyline fill="none" stroke="#000000" points="303,-107.5 512,-107.5 "/>
+<text text-anchor="start" x="311" y="-95.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ VisitAttrs()</text>
+<text text-anchor="start" x="311" y="-84.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TVM_DECLARE_FINAL_OBJECT_INFO()</text>
+<text text-anchor="start" x="311" y="-73.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
+<text text-anchor="start" x="311" y="-62.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Define()</text>
+<text text-anchor="start" x="311" y="-51.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetVarDoc()</text>
+<text text-anchor="start" x="311" y="-40.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsVarDefined()</text>
+<text text-anchor="start" x="311" y="-29.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RemoveVar()</text>
+<text text-anchor="start" x="311" y="-18.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ SetCommonPrefix()</text>
+<text text-anchor="start" x="311" y="-7.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ AsDoc()</text>
</g>
<!-- Node3 -->
<g id="node2" class="node">
@@ -77,8 +77,8 @@
<!-- Node3->Node2 -->
<g id="edge1" class="edge">
<title>Node3->Node2</title>
-<path fill="none" stroke="#191970" d="M190.0543,-308.7513C212.6571,-283.778 237.4456,-258.5677 262.5,-237 297.7277,-206.6747 339.9115,-177.9206 377.8747,-154.3205"/>
-<polygon fill="none" stroke="#191970" points="187.1845,-306.7089 183.1119,-316.4897 192.395,-311.3834 187.1845,-306.7089"/>
+<path fill="none" stroke="#191970" d="M189.9593,-316.4817C213.3705,-290.0439 238.5221,-262.2704 262.5,-237 277.2939,-221.4086 293.2704,-205.185 309.0259,-189.5158"/>
+<polygon fill="none" stroke="#191970" points="187.0782,-314.4567 183.0801,-324.2682 192.3241,-319.0914 187.0782,-314.4567"/>
</g>
<!-- Node3->Node3 -->
<g id="edge2" class="edge">
@@ -114,187 +114,149 @@
<!-- Node4->Node2 -->
<g id="edge3" class="edge">
<title>Node4->Node2</title>
-<path fill="none" stroke="#404040" d="M367.1735,-341.7923C379.5841,-308.566 394.4462,-270.7431 409.5,-237 414.7745,-225.1772 420.5541,-212.9422 426.4789,-200.8387"/>
-<polygon fill="none" stroke="#404040" points="426.6305,-200.5321 425.7051,-193.3806 431.9506,-189.7759 432.876,-196.9274 426.6305,-200.5321"/>
-<text text-anchor="middle" x="438.5" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +mod</text>
+<path fill="none" stroke="#404040" d="M354.6056,-341.8403C363.8414,-298.74 374.9958,-246.6864 384.6758,-201.5131"/>
+<polygon fill="none" stroke="#404040" points="384.7187,-201.3126 382.0647,-194.6076 387.2331,-189.579 389.8871,-196.2839 384.7187,-201.3126"/>
+<text text-anchor="middle" x="398.5" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +mod</text>
</g>
<!-- Node5 -->
<g id="node4" class="node">
<title>Node5</title>
<g id="a_node4"><a xlink:href="classtvm_1_1runtime_1_1ObjectRef.html" target="_top" xlink:title="Base class of all object reference. ">
-<polygon fill="#ffffff" stroke="#000000" points="490.5,-662.5 490.5,-884.5 624.5,-884.5 624.5,-662.5 490.5,-662.5"/>
-<text text-anchor="middle" x="557.5" y="-872.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectRef</text>
-<polyline fill="none" stroke="#000000" points="490.5,-865.5 624.5,-865.5 "/>
-<text text-anchor="start" x="498.5" y="-853.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
-<polyline fill="none" stroke="#000000" points="490.5,-846.5 624.5,-846.5 "/>
-<text text-anchor="start" x="498.5" y="-834.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
-<text text-anchor="start" x="498.5" y="-823.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
-<text text-anchor="start" x="498.5" y="-812.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ same_as()</text>
-<text text-anchor="start" x="498.5" y="-801.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator==()</text>
-<text text-anchor="start" x="498.5" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator!=()</text>
-<text text-anchor="start" x="498.5" y="-779.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator<()</text>
-<text text-anchor="start" x="498.5" y="-768.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined()</text>
-<text text-anchor="start" x="498.5" y="-757.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
-<text text-anchor="start" x="498.5" y="-746.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
-<text text-anchor="start" x="498.5" y="-735.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
-<text text-anchor="start" x="498.5" y="-724.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ use_count()</text>
-<text text-anchor="start" x="498.5" y="-713.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ as()</text>
-<text text-anchor="start" x="498.5" y="-702.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># get_mutable()</text>
-<text text-anchor="start" x="498.5" y="-691.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DowncastNoCheck()</text>
-<text text-anchor="start" x="498.5" y="-680.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># FFIClearAfterMove()</text>
-<text text-anchor="start" x="498.5" y="-669.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetDataPtr()</text>
+<polygon fill="#ffffff" stroke="#000000" points="419.5,-662.5 419.5,-884.5 553.5,-884.5 553.5,-662.5 419.5,-662.5"/>
+<text text-anchor="middle" x="486.5" y="-872.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectRef</text>
+<polyline fill="none" stroke="#000000" points="419.5,-865.5 553.5,-865.5 "/>
+<text text-anchor="start" x="427.5" y="-853.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_is_nullable</text>
+<polyline fill="none" stroke="#000000" points="419.5,-846.5 553.5,-846.5 "/>
+<text text-anchor="start" x="427.5" y="-834.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
+<text text-anchor="start" x="427.5" y="-823.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectRef()</text>
+<text text-anchor="start" x="427.5" y="-812.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ same_as()</text>
+<text text-anchor="start" x="427.5" y="-801.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator==()</text>
+<text text-anchor="start" x="427.5" y="-790.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator!=()</text>
+<text text-anchor="start" x="427.5" y="-779.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator<()</text>
+<text text-anchor="start" x="427.5" y="-768.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ defined()</text>
+<text text-anchor="start" x="427.5" y="-757.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
+<text text-anchor="start" x="427.5" y="-746.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
+<text text-anchor="start" x="427.5" y="-735.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
+<text text-anchor="start" x="427.5" y="-724.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ use_count()</text>
+<text text-anchor="start" x="427.5" y="-713.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ as()</text>
+<text text-anchor="start" x="427.5" y="-702.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># get_mutable()</text>
+<text text-anchor="start" x="427.5" y="-691.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DowncastNoCheck()</text>
+<text text-anchor="start" x="427.5" y="-680.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># FFIClearAfterMove()</text>
+<text text-anchor="start" x="427.5" y="-669.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetDataPtr()</text>
</a>
</g>
</g>
<!-- Node5->Node4 -->
<g id="edge4" class="edge">
<title>Node5->Node4</title>
-<path fill="none" stroke="#191970" d="M482.7569,-710.4068C456.9724,-685.7792 429.5593,-656.0173 409.5,-625 388.8479,-593.0659 372.6499,-554.5046 360.7423,-520.202"/>
-<polygon fill="none" stroke="#191970" points="480.6573,-713.2374 490.3364,-717.5462 485.4569,-708.1418 480.6573,-713.2374"/>
+<path fill="none" stroke="#191970" d="M422.6638,-653.1908C418.0815,-643.7136 413.6407,-634.238 409.5,-625 394.3872,-591.2833 379.5133,-553.4656 367.1098,-520.2362"/>
+<polygon fill="none" stroke="#191970" points="419.5637,-654.8186 427.0954,-662.27 425.8543,-651.7481 419.5637,-654.8186"/>
</g>
<!-- Node7 -->
<g id="node6" class="node">
<title>Node7</title>
-<g id="a_node6"><a xlink:href="classtvm_1_1runtime_1_1Map.html" target="_top" xlink:title="{tvm::runtime::Map\<\l tvm::runtime::String,\l tvm::runtime::String \>\n||+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ Map()\l+ operator=()\l+ operator=()\l+ at()\land 12 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="418,-336.5 418,-525.5 547,-525.5 547,-336.5 418,-336.5"/>
-<text text-anchor="start" x="426" y="-513.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Map<</text>
-<text text-anchor="start" x="426" y="-502.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> tvm::runtime::String,</text>
-<text text-anchor="middle" x="482.5" y="-491.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> tvm::runtime::String ></text>
-<polyline fill="none" stroke="#000000" points="418,-484.5 547,-484.5 "/>
-<text text-anchor="middle" x="482.5" y="-472.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="418,-465.5 547,-465.5 "/>
-<text text-anchor="start" x="426" y="-453.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-442.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-431.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-387.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Map()</text>
-<text text-anchor="start" x="426" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="426" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="426" y="-354.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ at()</text>
-<text text-anchor="start" x="426" y="-343.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 12 more...</text>
+<g id="a_node6"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::runtime::String \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
+<polygon fill="#ffffff" stroke="#000000" points="418,-342 418,-520 555,-520 555,-342 418,-342"/>
+<text text-anchor="start" x="426" y="-508" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
+<text text-anchor="middle" x="486.5" y="-497" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::String ></text>
+<polyline fill="none" stroke="#000000" points="418,-490 555,-490 "/>
+<text text-anchor="middle" x="486.5" y="-478" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="418,-471 555,-471 "/>
+<text text-anchor="start" x="426" y="-459" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-448" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-437" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-426" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-415" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-393" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-382" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="426" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="426" y="-360" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="426" y="-349" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
</a>
</g>
</g>
<!-- Node5->Node7 -->
<g id="edge7" class="edge">
<title>Node5->Node7</title>
-<path fill="none" stroke="#191970" d="M530.9765,-652.3761C521.8596,-610.742 511.7934,-564.7731 503.2074,-525.5637"/>
-<polygon fill="none" stroke="#191970" points="527.5915,-653.2802 533.1496,-662.3001 534.4295,-651.7828 527.5915,-653.2802"/>
+<path fill="none" stroke="#191970" d="M486.5,-652.2298C486.5,-608.7434 486.5,-560.5445 486.5,-520.2656"/>
+<polygon fill="none" stroke="#191970" points="483.0001,-652.3 486.5,-662.3001 490.0001,-652.3001 483.0001,-652.3"/>
</g>
<!-- Node8 -->
<g id="node7" class="node">
<title>Node8</title>
-<g id="a_node7"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::runtime::String \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="565,-342 565,-520 702,-520 702,-342 565,-342"/>
-<text text-anchor="start" x="573" y="-508" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
-<text text-anchor="middle" x="633.5" y="-497" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::String ></text>
-<polyline fill="none" stroke="#000000" points="565,-490 702,-490 "/>
-<text text-anchor="middle" x="633.5" y="-478" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="565,-471 702,-471 "/>
-<text text-anchor="start" x="573" y="-459" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-448" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-437" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-426" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-415" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-404" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-393" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-382" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="573" y="-371" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="573" y="-360" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="573" y="-349" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
+<g id="a_node7"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::script::printer\l::Frame \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
+<polygon fill="#ffffff" stroke="#000000" points="573,-336.5 573,-525.5 692,-525.5 692,-336.5 573,-336.5"/>
+<text text-anchor="start" x="581" y="-513.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
+<text text-anchor="start" x="581" y="-502.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
+<text text-anchor="middle" x="632.5" y="-491.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::Frame ></text>
+<polyline fill="none" stroke="#000000" points="573,-484.5 692,-484.5 "/>
+<text text-anchor="middle" x="632.5" y="-472.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="573,-465.5 692,-465.5 "/>
+<text text-anchor="start" x="581" y="-453.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-442.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-431.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-387.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
+<text text-anchor="start" x="581" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="581" y="-354.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="581" y="-343.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
</a>
</g>
</g>
<!-- Node5->Node8 -->
<g id="edge9" class="edge">
<title>Node5->Node8</title>
-<path fill="none" stroke="#191970" d="M584.3487,-652.5043C594.0141,-608.9464 604.7355,-560.6293 613.6922,-520.2656"/>
-<polygon fill="none" stroke="#191970" points="580.9245,-651.7793 582.175,-662.3001 587.7583,-653.2957 580.9245,-651.7793"/>
-</g>
-<!-- Node9 -->
-<g id="node8" class="node">
-<title>Node9</title>
-<g id="a_node8"><a xlink:href="classtvm_1_1runtime_1_1Array.html" target="_top" xlink:title="{tvm::runtime::Array\l\< tvm::script::printer\l::Frame \>\n||+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ Array()\l+ operator=()\l+ operator=()\land 25 more...\l}">
-<polygon fill="#ffffff" stroke="#000000" points="720,-336.5 720,-525.5 839,-525.5 839,-336.5 720,-336.5"/>
-<text text-anchor="start" x="728" y="-513.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Array</text>
-<text text-anchor="start" x="728" y="-502.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::script::printer</text>
-<text text-anchor="middle" x="779.5" y="-491.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::Frame ></text>
-<polyline fill="none" stroke="#000000" points="720,-484.5 839,-484.5 "/>
-<text text-anchor="middle" x="779.5" y="-472.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="720,-465.5 839,-465.5 "/>
-<text text-anchor="start" x="728" y="-453.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-442.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-431.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-387.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Array()</text>
-<text text-anchor="start" x="728" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="728" y="-354.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="728" y="-343.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 25 more...</text>
-</a>
-</g>
-</g>
-<!-- Node5->Node9 -->
-<g id="edge11" class="edge">
-<title>Node5->Node9</title>
-<path fill="none" stroke="#191970" d="M632.5985,-714.1239C660.4277,-689.0434 690.3283,-658.005 711.5,-625 730.77,-594.9597 745.3613,-558.7863 755.9887,-525.8552"/>
-<polygon fill="none" stroke="#191970" points="629.9671,-711.7798 624.8132,-721.0366 634.6148,-717.0143 629.9671,-711.7798"/>
+<path fill="none" stroke="#191970" d="M551.4884,-653.3489C556.0565,-643.8356 560.4499,-634.3086 564.5,-625 578.3118,-593.2556 591.4196,-557.7149 602.3739,-525.8124"/>
+<polygon fill="none" stroke="#191970" points="548.2845,-651.9342 547.0607,-662.4582 554.5802,-654.9944 548.2845,-651.9342"/>
</g>
<!-- Node6 -->
<g id="node5" class="node">
<title>Node6</title>
<g id="a_node5"><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="487.5,-932.5 487.5,-1110.5 627.5,-1110.5 627.5,-932.5 487.5,-932.5"/>
-<text text-anchor="start" x="495.5" y="-1098.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectPtr</text>
-<text text-anchor="middle" x="557.5" y="-1087.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::Object ></text>
-<polyline fill="none" stroke="#000000" points="487.5,-1080.5 627.5,-1080.5 "/>
-<text text-anchor="middle" x="557.5" y="-1068.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
-<polyline fill="none" stroke="#000000" points="487.5,-1061.5 627.5,-1061.5 "/>
-<text text-anchor="start" x="495.5" y="-1049.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-1038.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-1027.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-1016.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-1005.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-994.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-983.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~ObjectPtr()</text>
-<text text-anchor="start" x="495.5" y="-972.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ swap()</text>
-<text text-anchor="start" x="495.5" y="-961.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
-<text text-anchor="start" x="495.5" y="-950.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
-<text text-anchor="start" x="495.5" y="-939.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 11 more...</text>
+<polygon fill="#ffffff" stroke="#000000" points="416.5,-932.5 416.5,-1110.5 556.5,-1110.5 556.5,-932.5 416.5,-932.5"/>
+<text text-anchor="start" x="424.5" y="-1098.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::ObjectPtr</text>
+<text text-anchor="middle" x="486.5" y="-1087.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">< tvm::runtime::Object ></text>
+<polyline fill="none" stroke="#000000" points="416.5,-1080.5 556.5,-1080.5 "/>
+<text text-anchor="middle" x="486.5" y="-1068.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> </text>
+<polyline fill="none" stroke="#000000" points="416.5,-1061.5 556.5,-1061.5 "/>
+<text text-anchor="start" x="424.5" y="-1049.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-1038.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-1027.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-1016.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-1005.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-994.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-983.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ~ObjectPtr()</text>
+<text text-anchor="start" x="424.5" y="-972.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ swap()</text>
+<text text-anchor="start" x="424.5" y="-961.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ get()</text>
+<text text-anchor="start" x="424.5" y="-950.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator->()</text>
+<text text-anchor="start" x="424.5" y="-939.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">and 11 more...</text>
</a>
</g>
</g>
<!-- Node6->Node5 -->
<g id="edge5" class="edge">
<title>Node6->Node5</title>
-<path fill="none" stroke="#404040" d="M557.5,-932.3167C557.5,-920.8765 557.5,-909.0062 557.5,-897.1402"/>
-<polygon fill="none" stroke="#404040" points="557.5001,-896.7944 553.5,-890.7944 557.5,-884.7944 561.5,-890.7943 557.5001,-896.7944"/>
-<text text-anchor="middle" x="577" y="-906" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #data_</text>
+<path fill="none" stroke="#404040" d="M486.5,-932.3167C486.5,-920.8765 486.5,-909.0062 486.5,-897.1402"/>
+<polygon fill="none" stroke="#404040" points="486.5001,-896.7944 482.5,-890.7944 486.5,-884.7944 490.5,-890.7943 486.5001,-896.7944"/>
+<text text-anchor="middle" x="506" y="-906" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> #data_</text>
</g>
<!-- Node7->Node2 -->
<g id="edge6" class="edge">
<title>Node7->Node2</title>
-<path fill="none" stroke="#404040" d="M482.5,-336.0632C482.5,-294.2901 482.5,-245.0322 482.5,-201.9669"/>
-<polygon fill="none" stroke="#404040" points="482.5001,-201.7023 478.5,-195.7023 482.5,-189.7023 486.5,-195.7023 482.5001,-201.7023"/>
-<text text-anchor="middle" x="508.5" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +ir_prefix</text>
+<path fill="none" stroke="#404040" d="M465.5369,-341.8403C455.4032,-298.74 443.1644,-246.6864 432.5433,-201.5131"/>
+<polygon fill="none" stroke="#404040" points="432.4839,-201.2604 427.2168,-196.3352 429.7373,-189.579 435.0044,-194.5042 432.4839,-201.2604"/>
+<text text-anchor="middle" x="480" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +dispatch_tokens</text>
</g>
<!-- Node8->Node2 -->
<g id="edge8" class="edge">
<title>Node8->Node2</title>
-<path fill="none" stroke="#404040" d="M600.3158,-341.8264C587.3767,-308.604 571.9539,-270.7748 556.5,-237 551.098,-225.1939 545.2042,-212.9695 539.1777,-200.8722"/>
-<polygon fill="none" stroke="#404040" points="539.0079,-200.5344 532.7386,-196.9713 533.6163,-189.8138 539.8856,-193.3769 539.0079,-200.5344"/>
-<text text-anchor="middle" x="592" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +dispatch_tokens</text>
-</g>
-<!-- Node9->Node2 -->
-<g id="edge10" class="edge">
-<title>Node9->Node2</title>
-<path fill="none" stroke="#404040" d="M760.175,-336.1661C749.8645,-302.3106 734.4009,-265.5269 711.5,-237 681.3272,-199.4147 638.3483,-169.082 597.7587,-146.2521"/>
-<polygon fill="none" stroke="#404040" points="597.7392,-146.2414 590.5543,-146.8565 587.2232,-140.4609 594.408,-139.8458 597.7392,-146.2414"/>
-<text text-anchor="middle" x="714.5" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +frames</text>
+<path fill="none" stroke="#404040" d="M609.6788,-336.2919C599.0899,-303.1951 584.3461,-266.8773 564.5,-237 552.318,-218.6606 537.2401,-201.1284 521.2887,-185.03"/>
+<polygon fill="none" stroke="#404040" points="520.905,-184.6526 513.8224,-183.297 512.3497,-176.2378 519.4323,-177.5935 520.905,-184.6526"/>
+<text text-anchor="middle" x="573.5" y="-211" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"> +frames</text>
</g>
</g>
</svg>
diff --git a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg
index 1cc81a2fec..2227a6339b 100644
--- a/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg
+++ b/docs/reference/api/doxygen/classtvm_1_1script_1_1printer_1_1IRDocsifierNode__inherit__graph.svg
@@ -4,19 +4,18 @@
<!-- Generated by graphviz version 2.40.1 (20161225.0304)
-->
<!-- Title: tvm::script::printer::IRDocsifierNode Pages: 1 -->
-<svg width="217pt" height="677pt"
- viewBox="0.00 0.00 217.00 677.00" 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(4 673)">
+<svg width="217pt" height="666pt"
+ viewBox="0.00 0.00 217.00 666.00" 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(4 662)">
<title>tvm::script::printer::IRDocsifierNode</title>
-<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-673 213,-673 213,4 -4,4"/>
+<polygon fill="#ffffff" stroke="transparent" points="-4,4 -4,-662 213,-662 213,4 -4,4"/>
<!-- Node0 -->
<g id="node1" class="node">
<title>Node0</title>
-<polygon fill="#bfbfbf" stroke="#000000" points="0,-.5 0,-233.5 209,-233.5 209,-.5 0,-.5"/>
-<text text-anchor="start" x="8" y="-221.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
-<text text-anchor="middle" x="104.5" y="-210.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
-<polyline fill="none" stroke="#000000" points="0,-203.5 209,-203.5 "/>
-<text text-anchor="start" x="8" y="-191.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ ir_prefix</text>
+<polygon fill="#bfbfbf" stroke="#000000" points="0,-.5 0,-222.5 209,-222.5 209,-.5 0,-.5"/>
+<text text-anchor="start" x="8" y="-210.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::script::printer</text>
+<text text-anchor="middle" x="104.5" y="-199.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">::IRDocsifierNode</text>
+<polyline fill="none" stroke="#000000" points="0,-192.5 209,-192.5 "/>
<text text-anchor="start" x="8" y="-180.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ frames</text>
<text text-anchor="start" x="8" y="-169.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ dispatch_tokens</text>
<text text-anchor="start" x="8" y="-158.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ mod</text>
@@ -39,51 +38,51 @@
<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="13,-270.5 13,-668.5 196,-668.5 196,-270.5 13,-270.5"/>
-<text text-anchor="middle" x="104.5" y="-656.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
-<polyline fill="none" stroke="#000000" points="13,-649.5 196,-649.5 "/>
-<text text-anchor="start" x="21" y="-637.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
-<text text-anchor="start" x="21" y="-626.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_final</text>
-<text text-anchor="start" x="21" y="-615.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots</text>
-<text text-anchor="start" x="21" y="-604.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots_can</text>
-<text text-anchor="start" x="21" y="-593.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_overflow</text>
-<text text-anchor="start" x="21" y="-582.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_visit</text>
-<text text-anchor="start" x="21" y="-571.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_attrs</text>
-<text text-anchor="start" x="21" y="-560.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_sequal</text>
-<text text-anchor="start" x="21" y="-549.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="21" y="-538.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_shash</text>
-<text text-anchor="start" x="21" y="-527.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
-<text text-anchor="start" x="21" y="-516.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_index</text>
-<text text-anchor="start" x="21" y="-505.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># type_index_</text>
-<text text-anchor="start" x="21" y="-494.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># ref_counter_</text>
-<text text-anchor="start" x="21" y="-483.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># deleter_</text>
-<polyline fill="none" stroke="#000000" points="13,-476.5 196,-476.5 "/>
-<text text-anchor="start" x="21" y="-464.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_index()</text>
-<text text-anchor="start" x="21" y="-453.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKey()</text>
-<text text-anchor="start" x="21" y="-442.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKeyHash()</text>
-<text text-anchor="start" x="21" y="-431.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsInstance()</text>
-<text text-anchor="start" x="21" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
-<text text-anchor="start" x="21" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
+<polygon fill="#ffffff" stroke="#000000" points="13,-259.5 13,-657.5 196,-657.5 196,-259.5 13,-259.5"/>
+<text text-anchor="middle" x="104.5" y="-645.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">tvm::runtime::Object</text>
+<polyline fill="none" stroke="#000000" points="13,-638.5 196,-638.5 "/>
+<text text-anchor="start" x="21" y="-626.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_key</text>
+<text text-anchor="start" x="21" y="-615.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_final</text>
+<text text-anchor="start" x="21" y="-604.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots</text>
+<text text-anchor="start" x="21" y="-593.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_child_slots_can</text>
+<text text-anchor="start" x="21" y="-582.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_overflow</text>
+<text text-anchor="start" x="21" y="-571.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_visit</text>
+<text text-anchor="start" x="21" y="-560.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_attrs</text>
+<text text-anchor="start" x="21" y="-549.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_sequal</text>
+<text text-anchor="start" x="21" y="-538.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
+<text text-anchor="start" x="21" y="-527.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_has_method_shash</text>
+<text text-anchor="start" x="21" y="-516.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">_reduce</text>
+<text text-anchor="start" x="21" y="-505.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _type_index</text>
+<text text-anchor="start" x="21" y="-494.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># type_index_</text>
+<text text-anchor="start" x="21" y="-483.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># ref_counter_</text>
+<text text-anchor="start" x="21" y="-472.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># deleter_</text>
+<polyline fill="none" stroke="#000000" points="13,-465.5 196,-465.5 "/>
+<text text-anchor="start" x="21" y="-453.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ type_index()</text>
+<text text-anchor="start" x="21" y="-442.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKey()</text>
+<text text-anchor="start" x="21" y="-431.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ GetTypeKeyHash()</text>
+<text text-anchor="start" x="21" y="-420.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ IsInstance()</text>
+<text text-anchor="start" x="21" y="-409.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ unique()</text>
<text text-anchor="start" x="21" y="-398.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
<text text-anchor="start" x="21" y="-387.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
-<text text-anchor="start" x="21" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="21" y="-376.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ Object()</text>
<text text-anchor="start" x="21" y="-365.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
-<text text-anchor="start" x="21" y="-354.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2Key()</text>
-<text text-anchor="start" x="21" y="-343.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2KeyHash()</text>
-<text text-anchor="start" x="21" y="-332.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeKey2Index()</text>
-<text text-anchor="start" x="21" y="-321.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _GetOrAllocRuntimeTypeIndex()</text>
-<text text-anchor="start" x="21" y="-310.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RuntimeTypeIndex()</text>
-<text text-anchor="start" x="21" y="-299.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># IncRef()</text>
-<text text-anchor="start" x="21" y="-288.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DecRef()</text>
-<text text-anchor="start" x="21" y="-277.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetOrAllocRuntimeTypeIndex()</text>
+<text text-anchor="start" x="21" y="-354.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ operator=()</text>
+<text text-anchor="start" x="21" y="-343.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2Key()</text>
+<text text-anchor="start" x="21" y="-332.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeIndex2KeyHash()</text>
+<text text-anchor="start" x="21" y="-321.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ TypeKey2Index()</text>
+<text text-anchor="start" x="21" y="-310.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ _GetOrAllocRuntimeTypeIndex()</text>
+<text text-anchor="start" x="21" y="-299.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000">+ RuntimeTypeIndex()</text>
+<text text-anchor="start" x="21" y="-288.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># IncRef()</text>
+<text text-anchor="start" x="21" y="-277.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># DecRef()</text>
+<text text-anchor="start" x="21" y="-266.5" font-family="Helvetica,sans-Serif" font-size="10.00" fill="#000000"># GetOrAllocRuntimeTypeIndex()</text>
</a>
</g>
</g>
<!-- Node1->Node0 -->
<g id="edge1" class="edge">
<title>Node1->Node0</title>
-<path fill="none" stroke="#191970" d="M104.5,-260.1345C104.5,-251.1444 104.5,-242.2736 104.5,-233.6202"/>
-<polygon fill="none" stroke="#191970" points="101.0001,-260.3299 104.5,-270.33 108.0001,-260.33 101.0001,-260.3299"/>
+<path fill="none" stroke="#191970" d="M104.5,-249.1635C104.5,-240.1306 104.5,-231.2331 104.5,-222.575"/>
+<polygon fill="none" stroke="#191970" points="101.0001,-249.414 104.5,-259.4141 108.0001,-249.4141 101.0001,-249.414"/>
</g>
</g>
</svg>
diff --git a/docs/reference/api/doxygen/data__type_8h_source.html b/docs/reference/api/doxygen/data__type_8h_source.html
index bd9a91b7bf..347be9f6e7 100644
--- a/docs/reference/api/doxygen/data__type_8h_source.html
+++ b/docs/reference/api/doxygen/data__type_8h_source.html
@@ -66,7 +66,7 @@ $(function() {
<div class="title">data_type.h</div> </div>
</div><!--header-->
<div class="contents">
-<a href="data__type_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 more [...]
+<a href="data__type_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 more [...]
<div class="ttc" id="namespacetvm_1_1runtime_html_af22b89284299c81d0c1802199af446d7"><div class="ttname"><a href="namespacetvm_1_1runtime.html#af22b89284299c81d0c1802199af446d7">tvm::runtime::operator<<</a></div><div class="ttdeci">std::ostream & operator<<(std::ostream &os, const ObjectRef &n)</div><div class="ttdef"><b>Definition:</b> repr_printer.h:69</div></div>
<div class="ttc" id="namespacetvm_1_1runtime_html_a1b8eedadc5a269b2f9614ce69a9932e7"><div class="ttname"><a href="namespacetvm_1_1runtime.html#a1b8eedadc5a269b2f9614ce69a9932e7">tvm::runtime::GetCustomTypeRegistered</a></div><div class="ttdeci">bool GetCustomTypeRegistered(uint8_t type_code)</div><div class="ttdoc">Runtime utility for checking whether custom type is registered. </div></div>
<div class="ttc" id="classtvm_1_1runtime_1_1DataType_html_a4f3c849cf79c0812a50fdbb9ad175648"><div class="ttname"><a href="classtvm_1_1runtime_1_1DataType.html#a4f3c849cf79c0812a50fdbb9ad175648">tvm::runtime::DataType::is_int</a></div><div class="ttdeci">bool is_int() const</div><div class="ttdef"><b>Definition:</b> data_type.h:99</div></div>
diff --git a/docs/reference/api/doxygen/functions_func_i.html b/docs/reference/api/doxygen/functions_func_i.html
index 05274c911c..c780fe2077 100644
--- a/docs/reference/api/doxygen/functions_func_i.html
+++ b/docs/reference/api/doxygen/functions_func_i.html
@@ -263,7 +263,7 @@ $(function() {
: <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderFrame.html#a4ab7fb335968d6947f88b7b934c56354">tvm::script::ir_builder::IRBuilderFrame</a>
</li>
<li>IRDocsifier()
-: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a384e4e9461de539e28bb27efeefc9e19">tvm::script::printer::IRDocsifier</a>
+: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a495e99eb65db294cbc6afb33119266fd">tvm::script::printer::IRDocsifier</a>
</li>
<li>IRModule()
: <a class="el" href="classtvm_1_1IRModule.html#ad9109c0b0439b12e2e783267a11b5b7a">tvm::IRModule</a>
diff --git a/docs/reference/api/doxygen/functions_func_l.html b/docs/reference/api/doxygen/functions_func_l.html
index ca753170de..8f7b734cfa 100644
--- a/docs/reference/api/doxygen/functions_func_l.html
+++ b/docs/reference/api/doxygen/functions_func_l.html
@@ -191,7 +191,7 @@ $(function() {
: <a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunner.html#aa2b7cb576f9a7d5e900bc3159a07a7e9">tvm::auto_scheduler::LocalRunner</a>
</li>
<li>Lookup()
-: <a class="el" href="classtvm_1_1IRModuleNode.html#a8fb3c0a0e2c0a679753e6f8f45f54860">tvm::IRModuleNode</a>
+: <a class="el" href="classtvm_1_1IRModuleNode.html#a8430de337205d25778704d558f892ecb">tvm::IRModuleNode</a>
</li>
<li>LookupHashedValue()
: <a class="el" href="classtvm_1_1SHashHandlerDefault.html#a8d6338386dbf200ee7f73bee47c046b4">tvm::SHashHandlerDefault</a>
diff --git a/docs/reference/api/doxygen/functions_func_p.html b/docs/reference/api/doxygen/functions_func_p.html
index c16d6ebf30..3e8bb847eb 100644
--- a/docs/reference/api/doxygen/functions_func_p.html
+++ b/docs/reference/api/doxygen/functions_func_p.html
@@ -201,6 +201,9 @@ $(function() {
<li>PrefetchNode()
: <a class="el" href="classtvm_1_1tir_1_1PrefetchNode.html#acaaa5e89462c7edf3019df4283ec74db">tvm::tir::PrefetchNode</a>
</li>
+<li>Prefix()
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#abdc0ab77dcae93384cb88d8d123e6a7f">tvm::script::printer::Default</a>
+</li>
<li>PreloadMeasuredStates()
: <a class="el" href="classtvm_1_1auto__scheduler_1_1PreloadMeasuredStates.html#a67daf1ccd25a208fdf8d001f9a31d86b">tvm::auto_scheduler::PreloadMeasuredStates</a>
, <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchPolicyNode.html#abc2529d0b1cd485876e48037dd19dde1">tvm::auto_scheduler::SearchPolicyNode</a>
diff --git a/docs/reference/api/doxygen/functions_func_s.html b/docs/reference/api/doxygen/functions_func_s.html
index c1e31dfa27..0857b1a9d0 100644
--- a/docs/reference/api/doxygen/functions_func_s.html
+++ b/docs/reference/api/doxygen/functions_func_s.html
@@ -767,7 +767,7 @@ $(function() {
: <a class="el" href="classtvm_1_1runtime_1_1DeviceAPI.html#ac29b9295c432a87658392872c644864f">tvm::runtime::DeviceAPI</a>
</li>
<li>String()
-: <a class="el" href="classtvm_1_1runtime_1_1String.html#ac5d930b522e9fef9c07e51819d96d2f3">tvm::runtime::String</a>
+: <a class="el" href="classtvm_1_1runtime_1_1String.html#a68df7bab89fca339e3918438dd80300d">tvm::runtime::String</a>
</li>
<li>StringImm()
: <a class="el" href="classtvm_1_1tir_1_1StringImm.html#a0f2830290e055f677c5d5dea98aab726">tvm::tir::StringImm</a>
diff --git a/docs/reference/api/doxygen/functions_func_t.html b/docs/reference/api/doxygen/functions_func_t.html
index 89561c95d4..14300511e3 100644
--- a/docs/reference/api/doxygen/functions_func_t.html
+++ b/docs/reference/api/doxygen/functions_func_t.html
@@ -1234,10 +1234,10 @@ $(function() {
: <a class="el" href="classtvm_1_1TypeData.html#a0a98fd1095812379d2bd1337db1511c1">tvm::TypeData</a>
</li>
<li>TypedEnvFunc()
-: <a class="el" href="classtvm_1_1TypedEnvFunc_3_01R_07Args_8_8_8_08_4.html#a41a6b9014d0feeb628ca7edfd0d26f0b">tvm::TypedEnvFunc< R(Args...)></a>
+: <a class="el" href="classtvm_1_1TypedEnvFunc_3_01R_07Args_8_8_8_08_4.html#a0d72a6fa7263821c14bcd37837998ed9">tvm::TypedEnvFunc< R(Args...)></a>
</li>
<li>TypedPackedFunc()
-: <a class="el" href="classtvm_1_1runtime_1_1TypedPackedFunc_3_01R_07Args_8_8_8_08_4.html#a0161d426f9ca366c860ad48c384f7192">tvm::runtime::TypedPackedFunc< R(Args...)></a>
+: <a class="el" href="classtvm_1_1runtime_1_1TypedPackedFunc_3_01R_07Args_8_8_8_08_4.html#a8941c80982a1b2a289440f3c79bb0ac8">tvm::runtime::TypedPackedFunc< R(Args...)></a>
</li>
<li>TypeIndex2Key()
: <a class="el" href="classtvm_1_1runtime_1_1Object.html#a817ba6c23b7ee1821c48a75edf255a30">tvm::runtime::Object</a>
diff --git a/docs/reference/api/doxygen/functions_func_v.html b/docs/reference/api/doxygen/functions_func_v.html
index 153c745127..3353025590 100644
--- a/docs/reference/api/doxygen/functions_func_v.html
+++ b/docs/reference/api/doxygen/functions_func_v.html
@@ -89,6 +89,9 @@ $(function() {
<li>Vectorize()
: <a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html#ab4a8cd91959ceab22855ec338978bcee">tvm::tir::ScheduleNode</a>
</li>
+<li>VerboseExpr()
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#a2073f2f2ad1203129debf505af5af5c4">tvm::script::printer::Default</a>
+</li>
<li>VerifyGPUCode()
: <a class="el" href="classtvm_1_1meta__schedule_1_1Postproc.html#a7106b1742068c45966d6be5f4b8394aa">tvm::meta_schedule::Postproc</a>
</li>
@@ -429,17 +432,17 @@ $(function() {
, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#a6d35a6081ee7dbc440e5a980f70795c6">tvm::tir::StmtVisitor</a>
</li>
<li>VisitExpr_()
-: <a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html#a9259dfd39c812685890921178980e10a">tvm::relay::ExprFunctor< R(const Expr &n, Args...)></a>
-, <a class="el" href="classtvm_1_1relay_1_1ExprMutator.html#ae8e1f8512f277d0e2f0390ccad48b565">tvm::relay::ExprMutator</a>
-, <a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html#a52cec47d5e4792dd1cf0f5635ab14fa8">tvm::relay::ExprVisitor</a>
+: <a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html#aa49b8b27624631d2c2a637ec7ea5b98c">tvm::relay::ExprFunctor< R(const Expr &n, Args...)></a>
+, <a class="el" href="classtvm_1_1relay_1_1ExprMutator.html#ac22fb8d0001bb57b4dd4a796f32d4517">tvm::relay::ExprMutator</a>
+, <a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html#ab35bcfc19e95fbf444317fae7268f2fd">tvm::relay::ExprVisitor</a>
, <a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html#a86656f533b4961437f53d1dbe30ae1fb">tvm::relay::MixedModeMutator</a>
, <a class="el" href="classtvm_1_1relay_1_1MixedModeVisitor.html#a109b942e0299536851a3dc53a02b1ddc">tvm::relay::MixedModeVisitor</a>
-, <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#aa78adca6635007e175e6b17a4a758d40">tvm::tir::DataTypeLegalizer</a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html#a75f11907cca61167b90117b89e3b8818">tvm::tir::ExprFunctor< R(const PrimExpr &n, Args...)></a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprMutator.html#a9818e8f5c5e9e438e826b68adc0da578">tvm::tir::ExprMutator</a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html#ae8b5c68f5ad28542328676dca89dec5c">tvm::tir::ExprVisitor</a>
-, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeNormalizer.html#a305f82080e85e826fc70b30c635bb945">tvm::tir::IndexDataTypeNormalizer</a>
-, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#a046af3f6a45f44768b19a2d057bdca8f">tvm::tir::IndexDataTypeRewriter</a>
+, <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#a3bee1a8d9a9fd0ab7336810682919787">tvm::tir::DataTypeLegalizer</a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html#a70a0473f39ba12f4ca8f420c5cc2f28e">tvm::tir::ExprFunctor< R(const PrimExpr &n, Args...)></a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprMutator.html#a46b43325d9ad8cfe09818ba1dda4b46f">tvm::tir::ExprMutator</a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html#a001e91c6f7458afcc8f7433c5714dc38">tvm::tir::ExprVisitor</a>
+, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeNormalizer.html#a31ba40b84eda0ebf0c68bdacdbee78a8">tvm::tir::IndexDataTypeNormalizer</a>
+, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#a66c3c3fc73fa917b170922bff733d3a8">tvm::tir::IndexDataTypeRewriter</a>
</li>
<li>VisitExprDefault_()
: <a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html#ab35a37c57578e32a8c873cdfe9e31a0f">tvm::relay::ExprFunctor< R(const Expr &n, Args...)></a>
@@ -466,7 +469,7 @@ $(function() {
, <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#ad6692c86b749bb0d93042aa2a0425a74">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
</li>
<li>VisitPattern_()
-: <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#a11370205d1de851e817d40f031ad4811">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
+: <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#aa1bf3196b98aedddc028f14f5dcf5384">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
, <a class="el" href="classtvm_1_1relay_1_1PatternMutator.html#aedeb370baf4bca6018153d01d2594a84">tvm::relay::PatternMutator</a>
, <a class="el" href="classtvm_1_1relay_1_1PatternVisitor.html#ad5ed2a5c3b88ec027df9e4269dff4b80">tvm::relay::PatternVisitor</a>
</li>
@@ -484,11 +487,11 @@ $(function() {
, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#a4306d1beba05fa1ac582503498a6d7ce">tvm::tir::StmtMutator</a>
</li>
<li>VisitStmt_()
-: <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#aca20c14f1f058ffb3ef6c8bba1fda6f1">tvm::tir::DataTypeLegalizer</a>
-, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#abebbae492b191b36caae9663a4208532">tvm::tir::IndexDataTypeRewriter</a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#a087e79365d118d2de27a8616939e0d55">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#aecd16bf1a6715ea36f6c30e5dc2ceae7">tvm::tir::StmtMutator</a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#a3f0f6689f906e744cd40fd94e4aabed5">tvm::tir::StmtVisitor</a>
+: <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#a74a17d6246a62100e699dec154ceef72">tvm::tir::DataTypeLegalizer</a>
+, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#a8211e1dfd891b02917bd99531c8f22b2">tvm::tir::IndexDataTypeRewriter</a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#a5f86aa92770d279f28b47260f879b0bd">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#a758990a5b6adfb7585de67a20301ef0a">tvm::tir::StmtMutator</a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#a6cca1dd5ab6329207cecb8b94c63f8c2">tvm::tir::StmtVisitor</a>
</li>
<li>VisitStmtDefault_()
: <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#ae51b328e2b59a50bed7112a93dba1aae">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
@@ -502,8 +505,8 @@ $(function() {
, <a class="el" href="classtvm_1_1TypeMutator.html#a84e824911927d98e20a338eab8b75a45">tvm::TypeMutator</a>
</li>
<li>VisitType_()
-: <a class="el" href="classtvm_1_1TypeFunctor_3_01R_07const_01Type_01_6n_00_01Args_8_8_8_08_4.html#a40e3787a169f391bda4852aa7caf33cb">tvm::TypeFunctor< R(const Type &n, Args...)></a>
-, <a class="el" href="classtvm_1_1TypeMutator.html#a4c7667d35d0a9a28c957165b65536c93">tvm::TypeMutator</a>
+: <a class="el" href="classtvm_1_1TypeFunctor_3_01R_07const_01Type_01_6n_00_01Args_8_8_8_08_4.html#ae4061c0c04da5b6c889842506c2d5091">tvm::TypeFunctor< R(const Type &n, Args...)></a>
+, <a class="el" href="classtvm_1_1TypeMutator.html#a9c2d7e1a52faac66f55896ecde3f2211">tvm::TypeMutator</a>
, <a class="el" href="classtvm_1_1TypeVisitor.html#a3ed59cee793514b568fa0e9cb0ca437c">tvm::TypeVisitor</a>
</li>
<li>VisitTypeDefault_()
@@ -523,7 +526,7 @@ $(function() {
: <a class="el" href="structtvm_1_1runtime_1_1vm_1_1VMFrame.html#a8f8c990ee4fa7cb7472f5440f2ca3bde">tvm::runtime::vm::VMFrame</a>
</li>
<li>VMFunction()
-: <a class="el" href="structtvm_1_1runtime_1_1vm_1_1VMFunction.html#aea763069fe1dd6849ce0d1ec336931e0">tvm::runtime::vm::VMFunction</a>
+: <a class="el" href="structtvm_1_1runtime_1_1vm_1_1VMFunction.html#af9d2bdcf19642c21bc4909b9e9b6196d">tvm::runtime::vm::VMFunction</a>
</li>
<li>Void()
: <a class="el" href="classtvm_1_1runtime_1_1DataType.html#ab8dc0832aff8fd7421884c0fe20a3bfd">tvm::runtime::DataType</a>
diff --git a/docs/reference/api/doxygen/functions_i.html b/docs/reference/api/doxygen/functions_i.html
index 78aaa3be83..5d4d35572e 100644
--- a/docs/reference/api/doxygen/functions_i.html
+++ b/docs/reference/api/doxygen/functions_i.html
@@ -407,7 +407,7 @@ $(function() {
: <a class="el" href="structtvm_1_1relay_1_1ProposalAttrs.html#a0f8dbb0dfdc411a67254588e4ccf0f65">tvm::relay::ProposalAttrs</a>
</li>
<li>ir_prefix
-: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#adccf933a99d410b630f223ac1afe548f">tvm::script::printer::IRDocsifierNode</a>
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#ac124ee69e2c7ec4eb1d7743f337be24f">tvm::script::printer::Default</a>
</li>
<li>IRBuilder()
: <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilder.html#af63e228f3faaa0d87fdd52e1fe921dd3">tvm::script::ir_builder::IRBuilder</a>
@@ -416,7 +416,7 @@ $(function() {
: <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderFrame.html#a4ab7fb335968d6947f88b7b934c56354">tvm::script::ir_builder::IRBuilderFrame</a>
</li>
<li>IRDocsifier()
-: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a384e4e9461de539e28bb27efeefc9e19">tvm::script::printer::IRDocsifier</a>
+: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifier.html#a495e99eb65db294cbc6afb33119266fd">tvm::script::printer::IRDocsifier</a>
</li>
<li>irmod_storage_idx
: <a class="el" href="classtvm_1_1tir_1_1AllocateConstNode.html#a13983d7ae23271c7fed4794de0110c96">tvm::tir::AllocateConstNode</a>
diff --git a/docs/reference/api/doxygen/functions_l.html b/docs/reference/api/doxygen/functions_l.html
index fb213af109..cb5d79f984 100644
--- a/docs/reference/api/doxygen/functions_l.html
+++ b/docs/reference/api/doxygen/functions_l.html
@@ -282,7 +282,7 @@ $(function() {
, <a class="el" href="classtvm_1_1meta__schedule_1_1TuneContextNode.html#adb860776ff7f36d02596345ec0d09597">tvm::meta_schedule::TuneContextNode</a>
</li>
<li>Lookup()
-: <a class="el" href="classtvm_1_1IRModuleNode.html#a8430de337205d25778704d558f892ecb">tvm::IRModuleNode</a>
+: <a class="el" href="classtvm_1_1IRModuleNode.html#a8fb3c0a0e2c0a679753e6f8f45f54860">tvm::IRModuleNode</a>
</li>
<li>LookupHashedValue()
: <a class="el" href="classtvm_1_1SHashHandlerDefault.html#a8d6338386dbf200ee7f73bee47c046b4">tvm::SHashHandlerDefault</a>
@@ -292,7 +292,7 @@ $(function() {
: <a class="el" href="classtvm_1_1IRModuleNode.html#ae078ad8def39579701d144578c787bcf">tvm::IRModuleNode</a>
</li>
<li>LookupTypeDef()
-: <a class="el" href="classtvm_1_1IRModuleNode.html#ae095c1fd87642bd417224668c5b4d910">tvm::IRModuleNode</a>
+: <a class="el" href="classtvm_1_1IRModuleNode.html#a23f3769fe60b3b06c9d163650ea7caaf">tvm::IRModuleNode</a>
</li>
<li>loop_var
: <a class="el" href="classtvm_1_1tir_1_1ForNode.html#a7dbf66bdcf8ed397321517f0915a0946">tvm::tir::ForNode</a>
diff --git a/docs/reference/api/doxygen/functions_m.html b/docs/reference/api/doxygen/functions_m.html
index e62547588e..23a706ef21 100644
--- a/docs/reference/api/doxygen/functions_m.html
+++ b/docs/reference/api/doxygen/functions_m.html
@@ -344,7 +344,7 @@ $(function() {
: <a class="el" href="classtvm_1_1DiagnosticContextNode.html#adea7e38a6e47cbab7fb5639f208aa536">tvm::DiagnosticContextNode</a>
</li>
<li>Module()
-: <a class="el" href="classtvm_1_1runtime_1_1Module.html#abd1380b3f813c2b6acefca3aaef425f4">tvm::runtime::Module</a>
+: <a class="el" href="classtvm_1_1runtime_1_1Module.html#abfbc619b3b3166d63ec52e399c24bed9">tvm::runtime::Module</a>
, <a class="el" href="classtvm_1_1runtime_1_1ModuleNode.html#a21f639900c480510650969df9c74d17d">tvm::runtime::ModuleNode</a>
</li>
<li>module_handle
diff --git a/docs/reference/api/doxygen/functions_p.html b/docs/reference/api/doxygen/functions_p.html
index aee962b34b..982ce9d93b 100644
--- a/docs/reference/api/doxygen/functions_p.html
+++ b/docs/reference/api/doxygen/functions_p.html
@@ -350,7 +350,7 @@ $(function() {
: <a class="el" href="classtvm_1_1te_1_1IterVarAttrNode.html#aea7a6bc44a7ddca46c76c666eba37b7f">tvm::te::IterVarAttrNode</a>
</li>
<li>PragmaStep()
-: <a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStep.html#a9f3ec96f3e561a14d8d9235c4d46e2eb">tvm::auto_scheduler::PragmaStep</a>
+: <a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStep.html#a7692c2a9934af1f36b218840034a88d5">tvm::auto_scheduler::PragmaStep</a>
</li>
<li>pre_
: <a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html#a81d6c2593e361659ed2d0bea78a8f58a">tvm::relay::MixedModeMutator</a>
@@ -395,6 +395,9 @@ $(function() {
<li>PrefetchNode()
: <a class="el" href="classtvm_1_1tir_1_1PrefetchNode.html#a73ef244c364b9c7efaee36e6bec746e7">tvm::tir::PrefetchNode</a>
</li>
+<li>Prefix()
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#abdc0ab77dcae93384cb88d8d123e6a7f">tvm::script::printer::Default</a>
+</li>
<li>prefix_
: <a class="el" href="classtvm_1_1NameSupplyNode.html#aa14405ac3611e27389632477779fb6ad">tvm::NameSupplyNode</a>
</li>
diff --git a/docs/reference/api/doxygen/functions_s.html b/docs/reference/api/doxygen/functions_s.html
index b2ada6a275..37055f0b39 100644
--- a/docs/reference/api/doxygen/functions_s.html
+++ b/docs/reference/api/doxygen/functions_s.html
@@ -1077,7 +1077,7 @@ $(function() {
: <a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html#a93d1d23f24d903db844f75f51fe09a36">tvm::tir::ScheduleNode</a>
</li>
<li>StorageAlignStep()
-: <a class="el" href="classtvm_1_1auto__scheduler_1_1StorageAlignStep.html#af50b7c2f020f8e0a80f5bcc8e559b394">tvm::auto_scheduler::StorageAlignStep</a>
+: <a class="el" href="classtvm_1_1auto__scheduler_1_1StorageAlignStep.html#a99dbb8c55d9e7d78268b6d43fd348bc7">tvm::auto_scheduler::StorageAlignStep</a>
</li>
<li>StorageType
: <a class="el" href="classtvm_1_1runtime_1_1SimpleObjAllocator_1_1ArrayHandler.html#a67e86db3290b1d3bd4aca7e7a2faf187">tvm::runtime::SimpleObjAllocator::ArrayHandler< ArrayType, ElemType ></a>
@@ -1135,7 +1135,7 @@ $(function() {
, <a class="el" href="classtvm_1_1tir_1_1BufferNode.html#ac18ddd10b79a30ae57d3a8283686259d">tvm::tir::BufferNode</a>
</li>
<li>String()
-: <a class="el" href="classtvm_1_1runtime_1_1String.html#a02fca36e3ff55cc1e83635b02a11fca3">tvm::runtime::String</a>
+: <a class="el" href="classtvm_1_1runtime_1_1String.html#acf549b3c43142639879e0fc31ea5cd77">tvm::runtime::String</a>
, <a class="el" href="classtvm_1_1runtime_1_1StringObj_1_1FromStd.html#a7fb804f7dc96dd9f705c84095f37f1ca">tvm::runtime::StringObj::FromStd</a>
, <a class="el" href="classtvm_1_1runtime_1_1StringObj.html#a7fb804f7dc96dd9f705c84095f37f1ca">tvm::runtime::StringObj</a>
</li>
diff --git a/docs/reference/api/doxygen/functions_t.html b/docs/reference/api/doxygen/functions_t.html
index 297da45591..6a737ffccb 100644
--- a/docs/reference/api/doxygen/functions_t.html
+++ b/docs/reference/api/doxygen/functions_t.html
@@ -202,7 +202,7 @@ $(function() {
: <a class="el" href="classtvm_1_1te_1_1Stage.html#ab5fe485e1d730c36b096c060b8d2ef9d">tvm::te::Stage</a>
</li>
<li>Tensorize()
-: <a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html#a66983e2dde6aeb18b443616398fff8bf">tvm::tir::ScheduleNode</a>
+: <a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html#afa67abdb51145a49d42cd1464429d928">tvm::tir::ScheduleNode</a>
</li>
<li>TensorNode()
: <a class="el" href="classtvm_1_1te_1_1TensorNode.html#a153569448cb1bf9d2924d35639c3b8b8">tvm::te::TensorNode</a>
@@ -1428,7 +1428,7 @@ $(function() {
, <a class="el" href="classtvm_1_1runtime_1_1ObjectPtr.html#ae0ea8b4adc6dab8c74086bceaef6b3e1">tvm::runtime::ObjectPtr< T ></a>
, <a class="el" href="classtvm_1_1runtime_1_1ObjectRef.html#ae0ea8b4adc6dab8c74086bceaef6b3e1">tvm::runtime::ObjectRef</a>
, <a class="el" href="classtvm_1_1runtime_1_1TVMPODValue__.html#ae0ea8b4adc6dab8c74086bceaef6b3e1">tvm::runtime::TVMPODValue_</a>
-, <a class="el" href="classtvm_1_1runtime_1_1TVMRetValue.html#a77455a8fe7d27b90a01a64f1cd28e9ec">tvm::runtime::TVMRetValue</a>
+, <a class="el" href="classtvm_1_1runtime_1_1TVMRetValue.html#ac4a3850c0989e7c2d5cd8e0f096d0997">tvm::runtime::TVMRetValue</a>
, <a class="el" href="classtvm_1_1runtime_1_1TypedPackedFunc_3_01R_07Args_8_8_8_08_4.html#ae0ea8b4adc6dab8c74086bceaef6b3e1">tvm::runtime::TypedPackedFunc< R(Args...)></a>
</li>
<li>type
@@ -1500,7 +1500,7 @@ $(function() {
: <a class="el" href="classtvm_1_1TypedEnvFunc_3_01R_07Args_8_8_8_08_4.html#a41a6b9014d0feeb628ca7edfd0d26f0b">tvm::TypedEnvFunc< R(Args...)></a>
</li>
<li>TypedPackedFunc()
-: <a class="el" href="classtvm_1_1runtime_1_1TypedPackedFunc_3_01R_07Args_8_8_8_08_4.html#a0161d426f9ca366c860ad48c384f7192">tvm::runtime::TypedPackedFunc< R(Args...)></a>
+: <a class="el" href="classtvm_1_1runtime_1_1TypedPackedFunc_3_01R_07Args_8_8_8_08_4.html#af45a2ceff92e6f6c394ea766a45027a0">tvm::runtime::TypedPackedFunc< R(Args...)></a>
</li>
<li>TypeIndex2Key()
: <a class="el" href="classtvm_1_1runtime_1_1Object.html#a817ba6c23b7ee1821c48a75edf255a30">tvm::runtime::Object</a>
@@ -1523,7 +1523,7 @@ $(function() {
: <a class="el" href="classtvm_1_1TypeRelation.html#ac26b1897eab8197ed26606ab81b7403b">tvm::TypeRelation</a>
</li>
<li>TypeReporter()
-: <a class="el" href="classtvm_1_1TypeReporter.html#aa3dc38a3c84d324d0b3a9f358460a091">tvm::TypeReporter</a>
+: <a class="el" href="classtvm_1_1TypeReporter.html#a8e7e05a07f9f7ad9bea91f27afac9051">tvm::TypeReporter</a>
</li>
<li>types
: <a class="el" href="classtvm_1_1TupleAffineTypeNode.html#a30c834b7e1cb64467e6587ac16ebb187">tvm::TupleAffineTypeNode</a>
diff --git a/docs/reference/api/doxygen/functions_u.html b/docs/reference/api/doxygen/functions_u.html
index aee008c4c1..9051d7808e 100644
--- a/docs/reference/api/doxygen/functions_u.html
+++ b/docs/reference/api/doxygen/functions_u.html
@@ -122,7 +122,7 @@ $(function() {
, <a class="el" href="classtvm_1_1auto__scheduler_1_1CostModelNode.html#ae35b2b678760b8da57a43d3ae9c24da5">tvm::auto_scheduler::CostModelNode</a>
, <a class="el" href="classtvm_1_1auto__scheduler_1_1PythonBasedModelNode.html#a2d7849df6c7dbe93bf363c1d9f860a26">tvm::auto_scheduler::PythonBasedModelNode</a>
, <a class="el" href="classtvm_1_1auto__scheduler_1_1RandomModelNode.html#a7febac6c05d8e2d407f466467769ee32">tvm::auto_scheduler::RandomModelNode</a>
-, <a class="el" href="classtvm_1_1IRModuleNode.html#abdd8936c6fca33ef9b7c086f8fd58f84">tvm::IRModuleNode</a>
+, <a class="el" href="classtvm_1_1IRModuleNode.html#a94a93385e64ce844299729af6a573015">tvm::IRModuleNode</a>
, <a class="el" href="classtvm_1_1meta__schedule_1_1CostModelNode.html#a1bba32eba84db583fe90d1a5bce085f1">tvm::meta_schedule::CostModelNode</a>
, <a class="el" href="classtvm_1_1meta__schedule_1_1PyCostModelNode.html#a970b00b0eb1bf6b88eea2711b58c4d1d">tvm::meta_schedule::PyCostModelNode</a>
</li>
diff --git a/docs/reference/api/doxygen/functions_v.html b/docs/reference/api/doxygen/functions_v.html
index f354f11972..1a9c6ad53c 100644
--- a/docs/reference/api/doxygen/functions_v.html
+++ b/docs/reference/api/doxygen/functions_v.html
@@ -215,6 +215,12 @@ $(function() {
, <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchPolicyNode.html#a626eea16092d3c1a7d86497ee712f064">tvm::auto_scheduler::SearchPolicyNode</a>
, <a class="el" href="classtvm_1_1auto__scheduler_1_1TuningOptionsNode.html#ab5804969081fc4f1e1330b3e55da692f">tvm::auto_scheduler::TuningOptionsNode</a>
</li>
+<li>verbose_expr
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#ad5cabacf57ef0ad9b5a6287ef1e0df0f">tvm::script::printer::Default</a>
+</li>
+<li>VerboseExpr()
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#a2073f2f2ad1203129debf505af5af5c4">tvm::script::printer::Default</a>
+</li>
<li>VerifyGPUCode()
: <a class="el" href="classtvm_1_1meta__schedule_1_1Postproc.html#a7106b1742068c45966d6be5f4b8394aa">tvm::meta_schedule::Postproc</a>
</li>
@@ -564,8 +570,8 @@ $(function() {
, <a class="el" href="classtvm_1_1relay_1_1DFPatternVisitor.html#ae7e67d3a1709b0a180572417698ffaa8">tvm::relay::DFPatternVisitor</a>
</li>
<li>VisitDFPattern_()
-: <a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor_3_01R_07const_01DFPattern_01_6n_00_01Args_8_8_8_08_4.html#a4c61918b0ba22edf08576f16adb09a9d">tvm::relay::DFPatternFunctor< R(const DFPattern &n, Args...)></a>
-, <a class="el" href="classtvm_1_1relay_1_1DFPatternVisitor.html#a41da1b71513e124bc0b3dda3abafc5b3">tvm::relay::DFPatternVisitor</a>
+: <a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor_3_01R_07const_01DFPattern_01_6n_00_01Args_8_8_8_08_4.html#a7e2f9943f178774d1a54b52f92933a15">tvm::relay::DFPatternFunctor< R(const DFPattern &n, Args...)></a>
+, <a class="el" href="classtvm_1_1relay_1_1DFPatternVisitor.html#a5216eb09d409500f985d9b9044a1d69a">tvm::relay::DFPatternVisitor</a>
</li>
<li>VisitDFPatternDefault_()
: <a class="el" href="classtvm_1_1relay_1_1DFPatternFunctor_3_01R_07const_01DFPattern_01_6n_00_01Args_8_8_8_08_4.html#a5b505cf396e6efcd18aeacb0177eeb2a">tvm::relay::DFPatternFunctor< R(const DFPattern &n, Args...)></a>
@@ -586,15 +592,15 @@ $(function() {
, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#a6d35a6081ee7dbc440e5a980f70795c6">tvm::tir::StmtVisitor</a>
</li>
<li>VisitExpr_()
-: <a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html#ab067e6279e3e73817d41a17eef030726">tvm::relay::ExprFunctor< R(const Expr &n, Args...)></a>
-, <a class="el" href="classtvm_1_1relay_1_1ExprMutator.html#a568baaf4dd295369b950e7bcb248e32f">tvm::relay::ExprMutator</a>
-, <a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html#a8b1ef43d965026767385fb3ee5791928">tvm::relay::ExprVisitor</a>
-, <a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html#a7c3a1ed5bfb30ba993fdafa0d2d67fcb">tvm::relay::MixedModeMutator</a>
+: <a class="el" href="classtvm_1_1relay_1_1ExprFunctor_3_01R_07const_01Expr_01_6n_00_01Args_8_8_8_08_4.html#aa49b8b27624631d2c2a637ec7ea5b98c">tvm::relay::ExprFunctor< R(const Expr &n, Args...)></a>
+, <a class="el" href="classtvm_1_1relay_1_1ExprMutator.html#ac22fb8d0001bb57b4dd4a796f32d4517">tvm::relay::ExprMutator</a>
+, <a class="el" href="classtvm_1_1relay_1_1ExprVisitor.html#a72b3aeee22ea51f39c36d028d109cddc">tvm::relay::ExprVisitor</a>
+, <a class="el" href="classtvm_1_1relay_1_1MixedModeMutator.html#a86656f533b4961437f53d1dbe30ae1fb">tvm::relay::MixedModeMutator</a>
, <a class="el" href="classtvm_1_1relay_1_1MixedModeVisitor.html#a98c7381c097a64d67f47dd512d99f045">tvm::relay::MixedModeVisitor</a>
-, <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#ad6f857beadbb88b9db1c5bafaac59c8c">tvm::tir::DataTypeLegalizer</a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html#a55c70ed3207e5c48d5b8c0eae989cc41">tvm::tir::ExprFunctor< R(const PrimExpr &n, Args...)></a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprMutator.html#af6e722273818f70d0894aeecd1b55615">tvm::tir::ExprMutator</a>
-, <a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html#a0f5b65f0944eacc6401c8e02cb7ecd63">tvm::tir::ExprVisitor</a>
+, <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#aedf84c8c4df1f4c451d41d9c9b8fa266">tvm::tir::DataTypeLegalizer</a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprFunctor_3_01R_07const_01PrimExpr_01_6n_00_01Args_8_8_8_08_4.html#a46b99771934cdae3528e494d2e7f6c99">tvm::tir::ExprFunctor< R(const PrimExpr &n, Args...)></a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprMutator.html#ade08e1786bce8d1cb220d3f54f3fbfeb">tvm::tir::ExprMutator</a>
+, <a class="el" href="classtvm_1_1tir_1_1ExprVisitor.html#ae8b5c68f5ad28542328676dca89dec5c">tvm::tir::ExprVisitor</a>
, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeNormalizer.html#a83924f4f1bd318d388b492c6d6e5cd6b">tvm::tir::IndexDataTypeNormalizer</a>
, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#a9a63e5ef4cae7603732bda021380eb0c">tvm::tir::IndexDataTypeRewriter</a>
</li>
@@ -623,7 +629,7 @@ $(function() {
, <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#ad6692c86b749bb0d93042aa2a0425a74">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
</li>
<li>VisitPattern_()
-: <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#a11370205d1de851e817d40f031ad4811">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
+: <a class="el" href="classtvm_1_1relay_1_1PatternFunctor_3_01R_07const_01Pattern_01_6n_00_01Args_8_8_8_08_4.html#aa1bf3196b98aedddc028f14f5dcf5384">tvm::relay::PatternFunctor< R(const Pattern &n, Args...)></a>
, <a class="el" href="classtvm_1_1relay_1_1PatternMutator.html#a45f7cdfa9d72a3ab0ce2cb4ea04fec5b">tvm::relay::PatternMutator</a>
, <a class="el" href="classtvm_1_1relay_1_1PatternVisitor.html#aacc0309f0a28ade3e33287906c990259">tvm::relay::PatternVisitor</a>
</li>
@@ -641,11 +647,11 @@ $(function() {
, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#a4306d1beba05fa1ac582503498a6d7ce">tvm::tir::StmtMutator</a>
</li>
<li>VisitStmt_()
-: <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#a87e951cbdf97e52218f21db3fbbfaf38">tvm::tir::DataTypeLegalizer</a>
-, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#aeefb6cdc6b0619833d03167d8fd8e4c8">tvm::tir::IndexDataTypeRewriter</a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#a5943b89b1ee4b775e84d71eb43055f48">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#a758990a5b6adfb7585de67a20301ef0a">tvm::tir::StmtMutator</a>
-, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#aff2335e1aea1de67bdfb92271c8c0e10">tvm::tir::StmtVisitor</a>
+: <a class="el" href="classtvm_1_1tir_1_1DataTypeLegalizer.html#aca20c14f1f058ffb3ef6c8bba1fda6f1">tvm::tir::DataTypeLegalizer</a>
+, <a class="el" href="classtvm_1_1tir_1_1IndexDataTypeRewriter.html#ab0356b24b0e4d6887c86f1a2d647ac00">tvm::tir::IndexDataTypeRewriter</a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#afb4abf8cb69c4a9105eb38e262e96bc7">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtMutator.html#afda082108ac22091230cb937086e7a58">tvm::tir::StmtMutator</a>
+, <a class="el" href="classtvm_1_1tir_1_1StmtVisitor.html#a8e8cdf15d6e250ae57ae1777048e99f8">tvm::tir::StmtVisitor</a>
</li>
<li>VisitStmtDefault_()
: <a class="el" href="classtvm_1_1tir_1_1StmtFunctor_3_01R_07const_01Stmt_01_6n_00_01Args_8_8_8_01args_08_4.html#ae51b328e2b59a50bed7112a93dba1aae">tvm::tir::StmtFunctor< R(const Stmt &n, Args... args)></a>
@@ -659,9 +665,9 @@ $(function() {
, <a class="el" href="classtvm_1_1TypeMutator.html#a84e824911927d98e20a338eab8b75a45">tvm::TypeMutator</a>
</li>
<li>VisitType_()
-: <a class="el" href="classtvm_1_1TypeFunctor_3_01R_07const_01Type_01_6n_00_01Args_8_8_8_08_4.html#a05d03eb12860840ab5f28fa323f2beb4">tvm::TypeFunctor< R(const Type &n, Args...)></a>
-, <a class="el" href="classtvm_1_1TypeMutator.html#a18a04668d3fb464d957f3a26a4274104">tvm::TypeMutator</a>
-, <a class="el" href="classtvm_1_1TypeVisitor.html#a82c83b1524502579f56d194138badd3e">tvm::TypeVisitor</a>
+: <a class="el" href="classtvm_1_1TypeFunctor_3_01R_07const_01Type_01_6n_00_01Args_8_8_8_08_4.html#ae3a258acfcf5fe3ef0c7e291908d72ff">tvm::TypeFunctor< R(const Type &n, Args...)></a>
+, <a class="el" href="classtvm_1_1TypeMutator.html#a8171dc89a947d6224e83e86ce5d06d11">tvm::TypeMutator</a>
+, <a class="el" href="classtvm_1_1TypeVisitor.html#ac8845fbf58c1a1f0ebc23c7ee403aaab">tvm::TypeVisitor</a>
</li>
<li>VisitTypeDefault_()
: <a class="el" href="classtvm_1_1TypeFunctor_3_01R_07const_01Type_01_6n_00_01Args_8_8_8_08_4.html#a91553f9e04c39b3821a70ae4f7b0c597">tvm::TypeFunctor< R(const Type &n, Args...)></a>
diff --git a/docs/reference/api/doxygen/functions_vars_i.html b/docs/reference/api/doxygen/functions_vars_i.html
index 5ab877a7ae..c8f3fd4ad4 100644
--- a/docs/reference/api/doxygen/functions_vars_i.html
+++ b/docs/reference/api/doxygen/functions_vars_i.html
@@ -202,7 +202,7 @@ $(function() {
: <a class="el" href="structtvm_1_1relay_1_1ProposalAttrs.html#a0f8dbb0dfdc411a67254588e4ccf0f65">tvm::relay::ProposalAttrs</a>
</li>
<li>ir_prefix
-: <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html#adccf933a99d410b630f223ac1afe548f">tvm::script::printer::IRDocsifierNode</a>
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#ac124ee69e2c7ec4eb1d7743f337be24f">tvm::script::printer::Default</a>
</li>
<li>irmod_storage_idx
: <a class="el" href="classtvm_1_1tir_1_1AllocateConstNode.html#a13983d7ae23271c7fed4794de0110c96">tvm::tir::AllocateConstNode</a>
diff --git a/docs/reference/api/doxygen/functions_vars_v.html b/docs/reference/api/doxygen/functions_vars_v.html
index cb64ddf7de..a932ca77dc 100644
--- a/docs/reference/api/doxygen/functions_vars_v.html
+++ b/docs/reference/api/doxygen/functions_vars_v.html
@@ -168,6 +168,9 @@ $(function() {
, <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchPolicyNode.html#a626eea16092d3c1a7d86497ee712f064">tvm::auto_scheduler::SearchPolicyNode</a>
, <a class="el" href="classtvm_1_1auto__scheduler_1_1TuningOptionsNode.html#ab5804969081fc4f1e1330b3e55da692f">tvm::auto_scheduler::TuningOptionsNode</a>
</li>
+<li>verbose_expr
+: <a class="el" href="structtvm_1_1script_1_1printer_1_1Default.html#ad5cabacf57ef0ad9b5a6287ef1e0df0f">tvm::script::printer::Default</a>
+</li>
<li>version
: <a class="el" href="structTVMMetadata.html#ade3312efd4c0e5beaf390959621b2f52">TVMMetadata</a>
</li>
diff --git a/docs/reference/api/doxygen/hierarchy.html b/docs/reference/api/doxygen/hierarchy.html
index e48be4452f..364d25a30e 100644
--- a/docs/reference/api/doxygen/hierarchy.html
+++ b/docs/reference/api/doxygen/hierarchy.html
@@ -974,416 +974,415 @@ This inheritance list is sorted roughly, but not completely, alphabetically:</di
<tr id="row_113_68_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::runtime::String, tvm::GlobalVar ></a></td><td class="desc"></td></tr>
<tr id="row_113_69_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::runtime::String, tvm::runtime::Map< tvm::runtime::String, tvm::runtime::ObjectRef > ></a></td><td class="desc"></td></tr>
<tr id="row_113_70_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::runtime::String, tvm::runtime::ObjectRef ></a></td><td class="desc"></td></tr>
-<tr id="row_113_71_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::runtime::String, tvm::runtime::String ></a></td><td class="desc"></td></tr>
-<tr id="row_113_72_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::SourceName, tvm::parser::Source ></a></td><td class="desc"></td></tr>
-<tr id="row_113_73_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::Target, tvm::Integer ></a></td><td class="desc"></td></tr>
-<tr id="row_113_74_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::te::Operation, tvm::te::Stage ></a></td><td class="desc"></td></tr>
-<tr id="row_113_75_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Buffer, tvm::tir::Buffer ></a></td><td class="desc"></td></tr>
-<tr id="row_113_76_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Instruction, tvm::runtime::ObjectRef ></a></td><td class="desc"></td></tr>
-<tr id="row_113_77_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::IterVar, tvm::te::IterVarAttr ></a></td><td class="desc"></td></tr>
-<tr id="row_113_78_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::usmp::BufferInfo, tvm::tir::Stmt ></a></td><td class="desc"></td></tr>
-<tr id="row_113_79_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::PrimExpr ></a></td><td class="desc"></td></tr>
-<tr id="row_113_80_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::Range ></a></td><td class="desc"></td></tr>
-<tr id="row_113_81_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::tir::Buffer ></a></td><td class="desc"></td></tr>
-<tr id="row_113_82_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::tir::IterVar ></a></td><td class="desc"></td></tr>
-<tr id="row_113_83_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::DiagnosticContext ></a></td><td class="desc"></td></tr>
-<tr id="row_113_84_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::FloatImm ></a></td><td class="desc"></td></tr>
-<tr id="row_113_85_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Integer ></a></td><td class="desc"></td></tr>
-<tr id="row_113_86_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::IRModule ></a></td><td class="desc"></td></tr>
-<tr id="row_113_87_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::CostModel ></a></td><td class="desc"></td></tr>
-<tr id="row_113_88_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::Database ></a></td><td class="desc"></td></tr>
-<tr id="row_113_89_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::SearchStrategy ></a></td><td class="desc"></td></tr>
-<tr id="row_113_90_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::SpaceGenerator ></a></td><td class="desc"></td></tr>
-<tr id="row_113_91_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::PrimExpr ></a></td><td class="desc"></td></tr>
-<tr id="row_113_92_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::auto_scheduler::MeasureCallback > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_93_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::FloatImm > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_94_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::Integer > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_95_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::ArgInfo > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_96_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::BuilderResult > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_97_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::MeasureCandidate > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_98_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::Postproc > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_99_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::RunnerFuture > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_100_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::ScheduleRule > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_101_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::tir::BufferRegion > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_102_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::tir::Stmt > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_103_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::meta_schedule::Mutator, tvm::FloatImm > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_104_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::runtime::String, tvm::runtime::NDArray > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_105_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::runtime::String, tvm::runtime::ObjectRef > ></a></td><td class="desc"></td></tr>
-<tr id="row_113_106_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::NDArray ></a></td><td class="desc"></td></tr>
-<tr id="row_113_107_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::ObjectRef ></a></td><td class="desc"></td></tr>
-<tr id="row_113_108_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::String ></a></td><td class="desc"></td></tr>
-<tr id="row_113_109_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::script::printer::ExprDoc ></a></td><td class="desc"></td></tr>
-<tr id="row_113_110_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Target ></a></td><td class="desc"></td></tr>
-<tr id="row_113_111_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::tir::IterVar ></a></td><td class="desc"></td></tr>
-<tr id="row_113_112_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::tir::Stmt ></a></td><td class="desc"></td></tr>
-<tr id="row_113_113_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Type ></a></td><td class="desc"></td></tr>
-<tr id="row_113_114_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_114_" class="arrow" onclick="toggleFolder('113_114_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1AffineType.html" target="_self">tvm::AffineType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1AffineTypeNode.html" title="AffineType representation. ">A [...]
-<tr id="row_113_114_0_" 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="classtvm_1_1TensorAffineType.html" target="_self">tvm::TensorAffineType</a></td><td class="desc">Managed reference to AffineTypes </td></tr>
-<tr id="row_113_114_1_" 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="classtvm_1_1TupleAffineType.html" target="_self">tvm::TupleAffineType</a></td><td class="desc">Managed reference to TupleAffineTypes </td></tr>
-<tr id="row_113_115_" 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="classtvm_1_1arith_1_1ConstIntBound.html" target="_self">tvm::arith::ConstIntBound</a></td><td class="desc">Reference class to <a class="el" href="classtvm_1_1arith_1_1ConstIntBoundNode.html" title="Constant integer up and lower bound(inclusive). Useful for value bound analysis. ">Co [...]
-<tr id="row_113_116_" 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="classtvm_1_1arith_1_1IntConstraints.html" target="_self">tvm::arith::IntConstraints</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntConstraintsNode.html" title="Represent integer constrains including (integer) variables, their ranges and the [...]
-<tr id="row_113_117_" 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="classtvm_1_1arith_1_1IntConstraintsTransform.html" target="_self">tvm::arith::IntConstraintsTransform</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransformNode.html" title="We can have different set of variables to represent t [...]
-<tr id="row_113_118_" 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="classtvm_1_1arith_1_1IntGroupBounds.html" target="_self">tvm::arith::IntGroupBounds</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntGroupBoundsNode.html" title="Represent integer grouped bounds which are classified into lower bounds (inclusi [...]
-<tr id="row_113_119_" 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="classtvm_1_1arith_1_1IntSet.html" target="_self">tvm::arith::IntSet</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntSetNode.html" title="Base class of all Integer set containers. represent a set of integers in one dimension. ">IntSetNode</a> [...]
-<tr id="row_113_120_" 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="classtvm_1_1arith_1_1IterMapResult.html" target="_self">tvm::arith::IterMapResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterMapResultNode.html" title="Result of DetectIterMap. ">IterMapResultNode</a> </td></tr>
-<tr id="row_113_121_" 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="classtvm_1_1arith_1_1IterMark.html" target="_self">tvm::arith::IterMark</a></td><td class="desc">Managed reference to IterMarkExprNode </td></tr>
-<tr id="row_113_122_" 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="classtvm_1_1arith_1_1ModularSet.html" target="_self">tvm::arith::ModularSet</a></td><td class="desc">Reference of <a class="el" href="classtvm_1_1arith_1_1ModularSetNode.html" title="Range of a linear integer function. Use to do specify the possible index values. ">ModularSetNode</a [...]
-<tr id="row_113_123_" 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="classtvm_1_1AttrFieldInfo.html" target="_self">tvm::AttrFieldInfo</a></td><td class="desc"><a class="el" href="classtvm_1_1AttrFieldInfo.html" title="AttrFieldInfo. ">AttrFieldInfo</a> </td></tr>
-<tr id="row_113_124_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_124_" class="arrow" onclick="toggleFolder('113_124_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Attrs.html" target="_self">tvm::Attrs</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseAttrsNode.html" title="Base class of all attribute class. ">Base [...]
-<tr id="row_113_124_0_" 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="classtvm_1_1DictAttrs.html" target="_self">tvm::DictAttrs</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1DictAttrsNode.html" title="Specialized attribute type that is backed by a map. The DictAttrsNode implements the Attrs behavior...">DictAttrsNode< [...]
-<tr id="row_113_125_" 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="classtvm_1_1auto__scheduler_1_1AccessAnalyzer.html" target="_self">tvm::auto_scheduler::AccessAnalyzer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzerNode.html" title="Static analyzer for a ComputeDAG. ">AccessAnalyzerN [...]
-<tr id="row_113_126_" 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="classtvm_1_1auto__scheduler_1_1AttachMap.html" target="_self">tvm::auto_scheduler::AttachMap</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMapNode.html" title="stores the compute_at relation between stages This stores a bi-dir [...]
-<tr id="row_113_127_" 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="classtvm_1_1auto__scheduler_1_1BuildResult.html" target="_self">tvm::auto_scheduler::BuildResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResultNode.html" title="Store the result of a build. ">BuildResultNode</a> </td></tr>
-<tr id="row_113_128_" 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="classtvm_1_1auto__scheduler_1_1ComputeDAG.html" target="_self">tvm::auto_scheduler::ComputeDAG</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAGNode.html" title="The auto-scheduler's computational graph and related progra [...]
-<tr id="row_113_129_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_129_" class="arrow" onclick="toggleFolder('113_129_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModel.html" target="_self">tvm::auto_scheduler::CostModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CostMo [...]
-<tr id="row_113_129_0_" 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="classtvm_1_1auto__scheduler_1_1PythonBasedModel.html" target="_self">tvm::auto_scheduler::PythonBasedModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PythonBasedModelNode.html" title="A wrapper for cost model defined by python [...]
-<tr id="row_113_129_1_" 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="classtvm_1_1auto__scheduler_1_1RandomModel.html" target="_self">tvm::auto_scheduler::RandomModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RandomModelNode.html" title="The cost model returning random value for all predictions. [...]
-<tr id="row_113_130_" 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="classtvm_1_1auto__scheduler_1_1HardwareParams.html" target="_self">tvm::auto_scheduler::HardwareParams</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParamsNode.html" title="The parameters of target hardware used to guide the [...]
-<tr id="row_113_131_" 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="classtvm_1_1auto__scheduler_1_1Iterator.html" target="_self">tvm::auto_scheduler::Iterator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1IteratorNode.html" title="An iterator of a for-loop Similar to tvm::IterVar in include/tvm/tir/ [...]
-<tr id="row_113_132_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_132_" class="arrow" onclick="toggleFolder('113_132_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallback.html" target="_self">tvm::auto_scheduler::MeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__schedul [...]
-<tr id="row_113_132_0_" 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="classtvm_1_1auto__scheduler_1_1PythonBasedMeasureCallback.html" target="_self">tvm::auto_scheduler::PythonBasedMeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PythonBasedMeasureCallbackNode.html" title="A wrapper for [...]
-<tr id="row_113_132_1_" 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="classtvm_1_1auto__scheduler_1_1RecordToFile.html" target="_self">tvm::auto_scheduler::RecordToFile</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RecordToFileNode.html" title="Callback for logging the input and results of measureme [...]
-<tr id="row_113_133_" 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="classtvm_1_1auto__scheduler_1_1MeasureInput.html" target="_self">tvm::auto_scheduler::MeasureInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInputNode.html" title="Store the input of a measurement. ">MeasureInputNode</a> [...]
-<tr id="row_113_134_" 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="classtvm_1_1auto__scheduler_1_1MeasureResult.html" target="_self">tvm::auto_scheduler::MeasureResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureResultNode.html" title="Store the results of a measurement. ">MeasureResultNod [...]
-<tr id="row_113_135_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_135_" class="arrow" onclick="toggleFolder('113_135_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramBuilder.html" target="_self">tvm::auto_scheduler::ProgramBuilder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler [...]
-<tr id="row_113_135_0_" 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="classtvm_1_1auto__scheduler_1_1LocalBuilder.html" target="_self">tvm::auto_scheduler::LocalBuilder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1LocalBuilderNode.html" title="LocalBuilder use local CPU cores to build programs in p [...]
-<tr id="row_113_136_" 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="classtvm_1_1auto__scheduler_1_1ProgramMeasurer.html" target="_self">tvm::auto_scheduler::ProgramMeasurer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramMeasurerNode.html" title="Measurer that measures the time costs of tvm pro [...]
-<tr id="row_113_137_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_137_" class="arrow" onclick="toggleFolder('113_137_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramRunner.html" target="_self">tvm::auto_scheduler::ProgramRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1 [...]
-<tr id="row_113_137_0_" 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="classtvm_1_1auto__scheduler_1_1LocalRunner.html" target="_self">tvm::auto_scheduler::LocalRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunnerNode.html" title="LocalRunner that uses local CPU/GPU to measure the time co [...]
-<tr id="row_113_137_1_" 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="classtvm_1_1auto__scheduler_1_1RPCRunner.html" target="_self">tvm::auto_scheduler::RPCRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RPCRunnerNode.html" title="RPCRunner that uses RPC call to measures the time cost of progra [...]
-<tr id="row_113_138_" 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="classtvm_1_1auto__scheduler_1_1RecordReader.html" target="_self">tvm::auto_scheduler::RecordReader</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReaderNode.html" title="Log reader to load step logs from a file. ">RecordReaderN [...]
-<tr id="row_113_139_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_139_" class="arrow" onclick="toggleFolder('113_139_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1SearchCallback.html" target="_self">tvm::auto_scheduler::SearchCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler [...]
-<tr id="row_113_139_0_" 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="classtvm_1_1auto__scheduler_1_1PreloadMeasuredStates.html" target="_self">tvm::auto_scheduler::PreloadMeasuredStates</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PreloadMeasuredStatesNode.html" title="Preload measured states from [...]
-<tr id="row_113_140_" 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="classtvm_1_1auto__scheduler_1_1SearchPolicy.html" target="_self">tvm::auto_scheduler::SearchPolicy</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchPolicyNode.html" title="The base class of search policies. ">SearchPolicyNode</a> [...]
-<tr id="row_113_141_" 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="classtvm_1_1auto__scheduler_1_1SearchTask.html" target="_self">tvm::auto_scheduler::SearchTask</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchTaskNode.html" title="The computation information and hardware parameters for a speci [...]
-<tr id="row_113_142_" 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="classtvm_1_1auto__scheduler_1_1Stage.html" target="_self">tvm::auto_scheduler::Stage</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StageNode.html" title="A op stage in the compute declaration. Similar to te::Stage in include/tvm/te/ [...]
-<tr id="row_113_143_" 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="classtvm_1_1auto__scheduler_1_1State.html" target="_self">tvm::auto_scheduler::State</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StateNode.html" title="A state in the search process. It consists of the current loop structure and a [...]
-<tr id="row_113_144_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_144_" class="arrow" onclick="toggleFolder('113_144_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1Step.html" target="_self">tvm::auto_scheduler::Step</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StepNode.html" t [...]
-<tr id="row_113_144_0_" 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="classtvm_1_1auto__scheduler_1_1AnnotationStep.html" target="_self">tvm::auto_scheduler::AnnotationStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStepNode.html" title="Annotation step that corresponds to vectorize, pa [...]
-<tr id="row_113_144_1_" 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="classtvm_1_1auto__scheduler_1_1CacheReadStep.html" target="_self">tvm::auto_scheduler::CacheReadStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStepNode.html" title="Cache read step that corresponds to te::Schedule::ca [...]
-<tr id="row_113_144_2_" 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="classtvm_1_1auto__scheduler_1_1CacheWriteStep.html" target="_self">tvm::auto_scheduler::CacheWriteStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStepNode.html" title="Cache write step that corresponds to te::Schedule [...]
-<tr id="row_113_144_3_" 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="classtvm_1_1auto__scheduler_1_1ComputeAtStep.html" target="_self">tvm::auto_scheduler::ComputeAtStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStepNode.html" title="Compute at step that corresponds to te::Stage::compu [...]
-<tr id="row_113_144_4_" 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="classtvm_1_1auto__scheduler_1_1ComputeInlineStep.html" target="_self">tvm::auto_scheduler::ComputeInlineStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStepNode.html" title="Compute inline step that corresponds to [...]
-<tr id="row_113_144_5_" 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="classtvm_1_1auto__scheduler_1_1ComputeRootStep.html" target="_self">tvm::auto_scheduler::ComputeRootStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStepNode.html" title="Compute root step that corresponds to te::Stag [...]
-<tr id="row_113_144_6_" 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="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStep.html" target="_self">tvm::auto_scheduler::FollowFusedSplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStepNode.html" title="Similar to FollowSplitStep, but [...]
-<tr id="row_113_144_7_" 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="classtvm_1_1auto__scheduler_1_1FollowSplitStep.html" target="_self">tvm::auto_scheduler::FollowSplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStepNode.html" title="Similar to SplitStepNode, but uses split facto [...]
-<tr id="row_113_144_8_" 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="classtvm_1_1auto__scheduler_1_1FuseStep.html" target="_self">tvm::auto_scheduler::FuseStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStepNode.html" title="Fuse step that corresponds to te::Stage::fuse. ">FuseStepNode</a> < [...]
-<tr id="row_113_144_9_" 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="classtvm_1_1auto__scheduler_1_1PragmaStep.html" target="_self">tvm::auto_scheduler::PragmaStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStepNode.html" title="Pragma step that corresponds to te::Stage::pragma. ">PragmaSt [...]
-<tr id="row_113_144_10_" 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="classtvm_1_1auto__scheduler_1_1ReorderStep.html" target="_self">tvm::auto_scheduler::ReorderStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ReorderStepNode.html" title="Reorder step that corresponds to te::Stage::reorder. ">Re [...]
-<tr id="row_113_144_11_" 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="classtvm_1_1auto__scheduler_1_1RfactorStep.html" target="_self">tvm::auto_scheduler::RfactorStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RfactorStepNode.html" title="Reduction factor step that corresponds to te::Schedule::r [...]
-<tr id="row_113_144_12_" 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="classtvm_1_1auto__scheduler_1_1SplitStep.html" target="_self">tvm::auto_scheduler::SplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SplitStepNode.html" title="Split step that corresponds to te::Stage::split with additional [...]
-<tr id="row_113_144_13_" 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="classtvm_1_1auto__scheduler_1_1StorageAlignStep.html" target="_self">tvm::auto_scheduler::StorageAlignStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StorageAlignStepNode.html" title="Storage align step that corresponds to te: [...]
-<tr id="row_113_145_" 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="classtvm_1_1auto__scheduler_1_1TuningOptions.html" target="_self">tvm::auto_scheduler::TuningOptions</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1TuningOptionsNode.html" title="Tuning and measurement options. ">TuningOptionsNode</a [...]
-<tr id="row_113_146_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_146_" class="arrow" onclick="toggleFolder('113_146_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1BaseExpr.html" target="_self">tvm::BaseExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseExprNode.html" title="Base type of all the expressions. "> [...]
-<tr id="row_113_146_0_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_146_0_" class="arrow" onclick="toggleFolder('113_146_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1PrimExpr.html" target="_self">tvm::PrimExpr</a></td><td class="desc">Reference to <a class="el" href="classtvm_1_1PrimExprNode.html" title="Base node of all primitive expressions [...]
-<tr id="row_113_146_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_146_0_0_" class="arrow" onclick="toggleFolder('113_146_0_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterMapExpr.html" target="_self">tvm::arith::IterMapExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterMapExprNode.htm [...]
-<tr id="row_113_146_0_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterSplitExpr.html" target="_self">tvm::arith::IterSplitExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterSplitExprNode.html" title="Split of an iterator. ">IterSplitExprNode</a> </td></tr>
-<tr id="row_113_146_0_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterSumExpr.html" target="_self">tvm::arith::IterSumExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterSumExprNode.html" title="Fuse multiple iterators by summing them with scaling. ">IterSumExprNode</a> </td></tr>
-<tr id="row_113_146_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_1FloatImm.html" target="_self">tvm::FloatImm</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1FloatImmNode.html" title="Constant floating point literals in the program. ">FloatImmNode</a> </td></tr>
-<tr id="row_113_146_0_2_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_146_0_2_" class="arrow" onclick="toggleFolder('113_146_0_2_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1IntImm.html" target="_self">tvm::IntImm</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1IntImmNode.html" title="Constant integer literal [...]
-<tr id="row_113_146_0_2_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Bool.html" target="_self">tvm::Bool</a></td><td class="desc">Boolean constant </td></tr>
-<tr id="row_113_146_0_2_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Integer.html" target="_self">tvm::Integer</a></td><td class="desc">Container of constant int that adds more constructors </td></tr>
-<tr id="row_113_146_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_1tir_1_1Add.html" target="_self">tvm::tir::Add</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AddNode.html" title="a + b ">AddNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1And.html" target="_self">tvm::tir::And</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AndNode.html" title="a && b ">AndNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1Any.html" target="_self">tvm::tir::Any</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AnyNode.html" title="Any shape. ">AnyNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1Broadcast.html" target="_self">tvm::tir::Broadcast</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BroadcastNode.html" title="Create a vector where all the elements are value. ">BroadcastNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1BufferLoad.html" target="_self">tvm::tir::BufferLoad</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferLoadNode.html" title="Load value from the high dimension buffer. ">BufferLoadNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1Call.html" target="_self">tvm::tir::Call</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CallNode.html" title="Call node. ">CallNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1Cast.html" target="_self">tvm::tir::Cast</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CastNode.html" title="Cast value from one data type to another. ">CastNode</a> </td></tr>
-<tr id="row_113_146_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_1tir_1_1Div.html" target="_self">tvm::tir::Div</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DivNode.html" title="a / b in the C semnatics. ">DivNode</a> </td></tr>
-<tr id="row_113_146_0_11_" 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_1tir_1_1EQ.html" target="_self">tvm::tir::EQ</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1EQNode.html" title="a == b ">EQNode</a> </td></tr>
-<tr id="row_113_146_0_12_" 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_1tir_1_1FloorDiv.html" target="_self">tvm::tir::FloorDiv</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1FloorDivNode.html" title="Floor division, floor(a/b) ">FloorDivNode</a> </td></tr>
-<tr id="row_113_146_0_13_" 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_1tir_1_1FloorMod.html" target="_self">tvm::tir::FloorMod</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1FloorModNode.html" title="The remainder of the floordiv. ">FloorModNode</a> </td></tr>
-<tr id="row_113_146_0_14_" 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_1tir_1_1GE.html" target="_self">tvm::tir::GE</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1GENode.html" title="a >= b ">GENode</a> </td></tr>
-<tr id="row_113_146_0_15_" 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_1tir_1_1GT.html" target="_self">tvm::tir::GT</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1GTNode.html" title="a > b ">GTNode</a> </td></tr>
-<tr id="row_113_146_0_16_" 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_1tir_1_1LE.html" target="_self">tvm::tir::LE</a></td><td class="desc">Managed reference to <a class="el" href="structtvm_1_1tir_1_1LENode.html" title="a <= b ">LENode</a> </td></tr>
-<tr id="row_113_146_0_17_" 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_1tir_1_1Let.html" target="_self">tvm::tir::Let</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LetNode.html" title="Let binding. Bind var to value then evaluate body. ">LetNode</a> </td></tr>
-<tr id="row_113_146_0_18_" 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_1tir_1_1Load.html" target="_self">tvm::tir::Load</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LoadNode.html" title="Load the value from buffer_var. ">LoadNode</a> </td></tr>
-<tr id="row_113_146_0_19_" 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_1tir_1_1LT.html" target="_self">tvm::tir::LT</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LTNode.html" title="a < b ">LTNode</a> </td></tr>
-<tr id="row_113_146_0_20_" 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_1tir_1_1Max.html" target="_self">tvm::tir::Max</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MaxNode.html" title="max(a, b) ">MaxNode</a> </td></tr>
-<tr id="row_113_146_0_21_" 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_1tir_1_1Min.html" target="_self">tvm::tir::Min</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MinNode.html" title="min(a, b) ">MinNode</a> </td></tr>
-<tr id="row_113_146_0_22_" 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_1tir_1_1Mod.html" target="_self">tvm::tir::Mod</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ModNode.html" title="a % b in the C semnatics. ">ModNode</a> </td></tr>
-<tr id="row_113_146_0_23_" 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_1tir_1_1Mul.html" target="_self">tvm::tir::Mul</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MulNode.html" title="a * b ">MulNode</a> </td></tr>
-<tr id="row_113_146_0_24_" 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_1tir_1_1NE.html" target="_self">tvm::tir::NE</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1NENode.html" title="a != b ">NENode</a> </td></tr>
-<tr id="row_113_146_0_25_" 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_1tir_1_1Not.html" target="_self">tvm::tir::Not</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1NotNode.html" title="!a ">NotNode</a> </td></tr>
-<tr id="row_113_146_0_26_" 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_1tir_1_1Or.html" target="_self">tvm::tir::Or</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1OrNode.html" title="a || b ">OrNode</a> </td></tr>
-<tr id="row_113_146_0_27_" 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_1tir_1_1ProducerLoad.html" target="_self">tvm::tir::ProducerLoad</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ProducerLoadNode.html" title="Load value from the result produced by the producer. ">ProducerLoadNode</a> </td></tr>
-<tr id="row_113_146_0_28_" 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_1tir_1_1Ramp.html" target="_self">tvm::tir::Ramp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1RampNode.html" title="Construct a vector with lanes elements where its i-th element equals base + i * stride. This is useful to const [...]
-<tr id="row_113_146_0_29_" 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_1tir_1_1Reduce.html" target="_self">tvm::tir::Reduce</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ReduceNode.html" title="Reduction operator operator. ">ReduceNode</a> </td></tr>
-<tr id="row_113_146_0_30_" 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_1tir_1_1Select.html" target="_self">tvm::tir::Select</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1SelectNode.html" title="return true_value if condition is true, otherwise return false_value. ">SelectNode</a> </td></tr>
-<tr id="row_113_146_0_31_" 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_1tir_1_1Shuffle.html" target="_self">tvm::tir::Shuffle</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ShuffleNode.html" title="Shuffle instruction. vec = concat(vectors) result = (vec[indices[0]], vec[indices[1]] ...">ShuffleNode [...]
-<tr id="row_113_146_0_32_" 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_1tir_1_1StringImm.html" target="_self">tvm::tir::StringImm</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1StringImmNode.html" title="String constants, only used in asserts. ">StringImmNode</a> </td></tr>
-<tr id="row_113_146_0_33_" 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_1tir_1_1Sub.html" target="_self">tvm::tir::Sub</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1SubNode.html" title="a - b ">SubNode</a> </td></tr>
-<tr id="row_113_146_0_34_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_146_0_34_" class="arrow" onclick="toggleFolder('113_146_0_34_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1Var.html" target="_self">tvm::tir::Var</a></td><td class="desc">Named variable in TIR </td></tr>
-<tr id="row_113_146_0_34_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1SizeVar.html" target="_self">tvm::tir::SizeVar</a></td><td class="desc">Named variable represents a tensor index size </td></tr>
-<tr id="row_113_146_1_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_146_1_" class="arrow" onclick="toggleFolder('113_146_1_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1RelayExpr.html" target="_self">tvm::RelayExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1RelayExprNode.html" title="Base node of all non-primit [...]
-<tr id="row_113_146_1_0_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_146_1_0_" class="arrow" onclick="toggleFolder('113_146_1_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1BaseFunc.html" target="_self">tvm::BaseFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseFuncNode.html" title="Base node of all functio [...]
-<tr id="row_113_146_1_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1Function.html" target="_self">tvm::relay::Function</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1relay_1_1FunctionNode.html" title="Relay Function container. ">FunctionNode</a> </td></tr>
-<tr id="row_113_146_1_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1PrimFunc.html" target="_self">tvm::tir::PrimFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1PrimFuncNode.html" title="Primitive functions that contains TIR statements. ">PrimFuncNode</a> </td></tr>
-<tr id="row_113_146_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_1Constructor.html" target="_self">tvm::Constructor</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1ConstructorNode.html" title="ADT constructor. Constructors compare by pointer equality. ">ConstructorNode</a> </td></tr>
-<tr id="row_113_146_1_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_1GlobalVar.html" target="_self">tvm::GlobalVar</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1GlobalVarNode.html" title="Global variable that lives in the top-level module. ">GlobalVarNode</a> </td></tr>
-<tr id="row_113_146_1_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_1Op.html" target="_self">tvm::Op</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1OpNode.html" title="Primitive Op(builtin intrinsics) ">OpNode</a> </td></tr>
-<tr id="row_113_146_1_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_1relay_1_1Call.html" target="_self">tvm::relay::Call</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1Constant.html" target="_self">tvm::relay::Constant</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1If.html" target="_self">tvm::relay::If</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1Let.html" target="_self">tvm::relay::Let</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1Match.html" target="_self">tvm::relay::Match</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1RefCreate.html" target="_self">tvm::relay::RefCreate</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_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_1relay_1_1RefRead.html" target="_self">tvm::relay::RefRead</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_11_" 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_1relay_1_1RefWrite.html" target="_self">tvm::relay::RefWrite</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_12_" 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_1relay_1_1TempExpr.html" target="_self">tvm::relay::TempExpr</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_13_" 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_1relay_1_1Tuple.html" target="_self">tvm::relay::Tuple</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_14_" 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_1relay_1_1TupleGetItem.html" target="_self">tvm::relay::TupleGetItem</a></td><td class="desc"></td></tr>
-<tr id="row_113_146_1_15_" 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_1relay_1_1Var.html" target="_self">tvm::relay::Var</a></td><td class="desc"></td></tr>
-<tr id="row_113_147_" 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="classtvm_1_1CompilationConfig.html" target="_self">tvm::CompilationConfig</a></td><td class="desc">Managed reference class to <code><a class="el" href="classtvm_1_1CompilationConfig.html" title="Managed reference class to CompilationConfig. ">CompilationConfig</a></code> </td></tr>
-<tr id="row_113_148_" 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="classtvm_1_1ConstantInfo.html" target="_self">tvm::ConstantInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_149_" 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="classtvm_1_1ConstantMemoryPools.html" target="_self">tvm::ConstantMemoryPools</a></td><td class="desc"></td></tr>
-<tr id="row_113_150_" 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="classtvm_1_1Diagnostic.html" target="_self">tvm::Diagnostic</a></td><td class="desc"></td></tr>
-<tr id="row_113_151_" 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="classtvm_1_1DiagnosticContext.html" target="_self">tvm::DiagnosticContext</a></td><td class="desc"></td></tr>
-<tr id="row_113_152_" 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="classtvm_1_1DiagnosticRenderer.html" target="_self">tvm::DiagnosticRenderer</a></td><td class="desc"></td></tr>
-<tr id="row_113_153_" 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="classtvm_1_1EnvFunc.html" target="_self">tvm::EnvFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1EnvFuncNode.html" title="A serializable function backed by TVM's global environment. ">EnvFuncNode</a> </td></tr>
-<tr id="row_113_154_" 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="classtvm_1_1GenericFunc.html" target="_self">tvm::GenericFunc</a></td><td class="desc">Generic function that can be specialized on a per-target basis </td></tr>
-<tr id="row_113_155_" 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="classtvm_1_1GlobalVarSupply.html" target="_self">tvm::GlobalVarSupply</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1GlobalVarSupplyNode.html" title="GlobalVarSupply can be used to generate unique GlobalVars. ">GlobalVarSupplyNode</a> </td></tr>
-<tr id="row_113_156_" 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="classtvm_1_1instrument_1_1PassInstrument.html" target="_self">tvm::instrument::PassInstrument</a></td><td class="desc">Managed reference class for <a class="el" href="classtvm_1_1instrument_1_1PassInstrumentNode.html" title="PassInstrumentNode forms an instrument implementation. It [...]
-<tr id="row_113_157_" 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="classtvm_1_1IRModule.html" target="_self">tvm::IRModule</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1IRModuleNode.html" title="IRModule that holds functions and type definitions. ">IRModuleNode</a> </td></tr>
-<tr id="row_113_158_" 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="classtvm_1_1MemoryInfo.html" target="_self">tvm::MemoryInfo</a></td><td class="desc">Defines memory info </td></tr>
-<tr id="row_113_159_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_159_" class="arrow" onclick="toggleFolder('113_159_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfo.html" target="_self">tvm::meta_schedule::ArgInfo</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfoNode.h [...]
-<tr id="row_113_159_0_" 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="classtvm_1_1meta__schedule_1_1TensorInfo.html" target="_self">tvm::meta_schedule::TensorInfo</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TensorInfoNode.html" title="The tensor argument information. ">TensorInfoNode</a> </td></tr>
-<tr id="row_113_160_" 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="classtvm_1_1meta__schedule_1_1Builder.html" target="_self">tvm::meta_schedule::Builder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderNode.html" title="The abstract builder interface. ">BuilderNode</a> </td></tr>
-<tr id="row_113_161_" 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="classtvm_1_1meta__schedule_1_1BuilderInput.html" target="_self">tvm::meta_schedule::BuilderInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInputNode.html" title="The builder's input, containing an IRModule and the targ [...]
-<tr id="row_113_162_" 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="classtvm_1_1meta__schedule_1_1BuilderResult.html" target="_self">tvm::meta_schedule::BuilderResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResultNode.html" title="The builder's output, containing the artifact path or [...]
-<tr id="row_113_163_" 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="classtvm_1_1meta__schedule_1_1CostModel.html" target="_self">tvm::meta_schedule::CostModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1CostModelNode.html" title="Cost model. ">CostModelNode</a> </td></tr>
-<tr id="row_113_164_" 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="classtvm_1_1meta__schedule_1_1Database.html" target="_self">tvm::meta_schedule::Database</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1DatabaseNode.html">DatabaseNode</a> </td></tr>
-<tr id="row_113_165_" 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="classtvm_1_1meta__schedule_1_1ExtractedTask.html" target="_self">tvm::meta_schedule::ExtractedTask</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTaskNode.html" title="A tuning task extracted from the high-level IR. ">Extract [...]
-<tr id="row_113_166_" 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="classtvm_1_1meta__schedule_1_1FeatureExtractor.html" target="_self">tvm::meta_schedule::FeatureExtractor</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractorNode.html" title="Extractor for features from measure candidates f [...]
-<tr id="row_113_167_" 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="classtvm_1_1meta__schedule_1_1MeasureCallback.html" target="_self">tvm::meta_schedule::MeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallbackNode.html" title="Rules to apply after measure results is available. [...]
-<tr id="row_113_168_" 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="classtvm_1_1meta__schedule_1_1MeasureCandidate.html" target="_self">tvm::meta_schedule::MeasureCandidate</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidateNode.html" title="The schedule (with input shapes) to be measured. [...]
-<tr id="row_113_169_" 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="classtvm_1_1meta__schedule_1_1Mutator.html" target="_self">tvm::meta_schedule::Mutator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MutatorNode.html" title="Mutator is designed to mutate the trace to explore the design space. ">Muta [...]
-<tr id="row_113_170_" 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="classtvm_1_1meta__schedule_1_1Postproc.html" target="_self">tvm::meta_schedule::Postproc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1PostprocNode.html" title="Rules to apply a postprocessor to a schedule. ">PostprocNode</a> </td></tr>
-<tr id="row_113_171_" 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="classtvm_1_1meta__schedule_1_1Profiler.html" target="_self">tvm::meta_schedule::Profiler</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ProfilerNode.html" title="A generic profiler. ">ProfilerNode</a> </td></tr>
-<tr id="row_113_172_" 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="classtvm_1_1meta__schedule_1_1Runner.html" target="_self">tvm::meta_schedule::Runner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerNode.html" title="The abstract runner interface. ">RunnerNode</a> </td></tr>
-<tr id="row_113_173_" 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="classtvm_1_1meta__schedule_1_1RunnerFuture.html" target="_self">tvm::meta_schedule::RunnerFuture</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerFutureNode.html" title="A class to asynchronously fetch runner's output. ">Runne [...]
-<tr id="row_113_174_" 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="classtvm_1_1meta__schedule_1_1RunnerInput.html" target="_self">tvm::meta_schedule::RunnerInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerInputNode.html" title="Runner's input containing path of artifact, type of device [...]
-<tr id="row_113_175_" 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="classtvm_1_1meta__schedule_1_1RunnerResult.html" target="_self">tvm::meta_schedule::RunnerResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerResultNode.html" title="Runner's output containing measurement result of Measure [...]
-<tr id="row_113_176_" 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="classtvm_1_1meta__schedule_1_1ScheduleRule.html" target="_self">tvm::meta_schedule::ScheduleRule</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ScheduleRuleNode.html" title="Rules to modify a block in a schedule. ">ScheduleRuleNode</a [...]
-<tr id="row_113_177_" 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="classtvm_1_1meta__schedule_1_1SearchStrategy.html" target="_self">tvm::meta_schedule::SearchStrategy</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1SearchStrategyNode.html" title="The search strategy for measure candidates generation. [...]
-<tr id="row_113_178_" 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="classtvm_1_1meta__schedule_1_1SpaceGenerator.html" target="_self">tvm::meta_schedule::SpaceGenerator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1SpaceGeneratorNode.html" title="The abstract class for design space generation. ">Spac [...]
-<tr id="row_113_179_" 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="classtvm_1_1meta__schedule_1_1TaskRecord.html" target="_self">tvm::meta_schedule::TaskRecord</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TaskRecordNode.html">TaskRecordNode</a> </td></tr>
-<tr id="row_113_180_" 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="classtvm_1_1meta__schedule_1_1TaskScheduler.html" target="_self">tvm::meta_schedule::TaskScheduler</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TaskSchedulerNode.html" title="The abstract interface of task schedulers. ">TaskSchedule [...]
-<tr id="row_113_181_" 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="classtvm_1_1meta__schedule_1_1TuneContext.html" target="_self">tvm::meta_schedule::TuneContext</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TuneContextNode.html" title="The auto tuning context. ">TuneContextNode</a> </td></tr>
-<tr id="row_113_182_" 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="classtvm_1_1meta__schedule_1_1TuningRecord.html" target="_self">tvm::meta_schedule::TuningRecord</a></td><td class="desc">The managed reference of <a class="el" href="classtvm_1_1meta__schedule_1_1TuningRecordNode.html" title="The class of tuning records. ">TuningRecordNode</a> </td></tr>
-<tr id="row_113_183_" 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="classtvm_1_1meta__schedule_1_1Workload.html" target="_self">tvm::meta_schedule::Workload</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1WorkloadNode.html" title="A workload, i.e. an IRModule and its structural hash. ">WorkloadNode</a> [...]
-<tr id="row_113_184_" 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="classtvm_1_1NameSupply.html" target="_self">tvm::NameSupply</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1NameSupplyNode.html" title="NameSupply can be used to generate unique names. ">NameSupplyNode</a> </td></tr>
-<tr id="row_113_185_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_185_" class="arrow" onclick="toggleFolder('113_185_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1ObjectPath.html" target="_self">tvm::ObjectPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_0_" 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="classtvm_1_1ArrayIndexPath.html" target="_self">tvm::ArrayIndexPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_1_" 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="classtvm_1_1AttributeAccessPath.html" target="_self">tvm::AttributeAccessPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_2_" 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="classtvm_1_1MapValuePath.html" target="_self">tvm::MapValuePath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_3_" 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="classtvm_1_1MissingArrayElementPath.html" target="_self">tvm::MissingArrayElementPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_4_" 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="classtvm_1_1MissingMapEntryPath.html" target="_self">tvm::MissingMapEntryPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_5_" 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="classtvm_1_1RootPath.html" target="_self">tvm::RootPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_185_6_" 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="classtvm_1_1UnknownAttributeAccessPath.html" target="_self">tvm::UnknownAttributeAccessPath</a></td><td class="desc"></td></tr>
-<tr id="row_113_186_" 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="classtvm_1_1ObjectPathPair.html" target="_self">tvm::ObjectPathPair</a></td><td class="desc"></td></tr>
-<tr id="row_113_187_" 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="classtvm_1_1parser_1_1Source.html" target="_self">tvm::parser::Source</a></td><td class="desc"></td></tr>
-<tr id="row_113_188_" 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="classtvm_1_1parser_1_1SourceMap.html" target="_self">tvm::parser::SourceMap</a></td><td class="desc"></td></tr>
-<tr id="row_113_189_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_189_" class="arrow" onclick="toggleFolder('113_189_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1PoolInfo.html" target="_self">tvm::PoolInfo</a></td><td class="desc">Base class for <a class="el" href="classtvm_1_1WorkspacePoolInfo.html">WorkspacePoolInfo</a> and <a class="el" href [...]
-<tr id="row_113_189_0_" 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="classtvm_1_1ConstantPoolInfo.html" target="_self">tvm::ConstantPoolInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_189_1_" 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="classtvm_1_1WorkspacePoolInfo.html" target="_self">tvm::WorkspacePoolInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_190_" 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="classtvm_1_1PoolInfoProperties.html" target="_self">tvm::PoolInfoProperties</a></td><td class="desc"></td></tr>
-<tr id="row_113_191_" 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="classtvm_1_1Range.html" target="_self">tvm::Range</a></td><td class="desc"><a class="el" href="classtvm_1_1Range.html" title="Range constainer. ">Range</a> constainer </td></tr>
-<tr id="row_113_192_" 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="classtvm_1_1relay_1_1Clause.html" target="_self">tvm::relay::Clause</a></td><td class="desc"></td></tr>
-<tr id="row_113_193_" 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="classtvm_1_1relay_1_1ConstructorValue.html" target="_self">tvm::relay::ConstructorValue</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_194_" class="arrow" onclick="toggleFolder('113_194_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1DFPattern.html" target="_self">tvm::relay::DFPattern</a></td><td class="desc">Managed reference to dataflow patterns </td></tr>
-<tr id="row_113_194_0_" 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="classtvm_1_1relay_1_1AltPattern.html" target="_self">tvm::relay::AltPattern</a></td><td class="desc">A pattern which matches either of two patterns </td></tr>
-<tr id="row_113_194_1_" 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="classtvm_1_1relay_1_1AttrPattern.html" target="_self">tvm::relay::AttrPattern</a></td><td class="desc">A pattern which matches attributes in another pattern </td></tr>
-<tr id="row_113_194_2_" 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="classtvm_1_1relay_1_1CallPattern.html" target="_self">tvm::relay::CallPattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_3_" 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="classtvm_1_1relay_1_1ConstantPattern.html" target="_self">tvm::relay::ConstantPattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_4_" 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="classtvm_1_1relay_1_1DataTypePattern.html" target="_self">tvm::relay::DataTypePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
-<tr id="row_113_194_5_" 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="classtvm_1_1relay_1_1DominatorPattern.html" target="_self">tvm::relay::DominatorPattern</a></td><td class="desc">A pattern which matches a variable length dominator path </td></tr>
-<tr id="row_113_194_6_" 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="classtvm_1_1relay_1_1ExprPattern.html" target="_self">tvm::relay::ExprPattern</a></td><td class="desc">A pattern which matches a literal expression </td></tr>
-<tr id="row_113_194_7_" 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="classtvm_1_1relay_1_1FunctionPattern.html" target="_self">tvm::relay::FunctionPattern</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1relay_1_1FunctionNode.html" title="Relay Function container. ">FunctionNode</a> </td></tr>
-<tr id="row_113_194_8_" 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="classtvm_1_1relay_1_1IfPattern.html" target="_self">tvm::relay::IfPattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_9_" 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="classtvm_1_1relay_1_1LetPattern.html" target="_self">tvm::relay::LetPattern</a></td><td class="desc"><a class="el" href="classtvm_1_1relay_1_1Let.html">Let</a> binding that binds a local var </td></tr>
-<tr id="row_113_194_10_" 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="classtvm_1_1relay_1_1ShapePattern.html" target="_self">tvm::relay::ShapePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
-<tr id="row_113_194_11_" 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="classtvm_1_1relay_1_1TupleGetItemPattern.html" target="_self">tvm::relay::TupleGetItemPattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_12_" 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="classtvm_1_1relay_1_1TuplePattern.html" target="_self">tvm::relay::TuplePattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_13_" 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="classtvm_1_1relay_1_1TypePattern.html" target="_self">tvm::relay::TypePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
-<tr id="row_113_194_14_" 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="classtvm_1_1relay_1_1VarPattern.html" target="_self">tvm::relay::VarPattern</a></td><td class="desc"></td></tr>
-<tr id="row_113_194_15_" 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="classtvm_1_1relay_1_1WildcardPattern.html" target="_self">tvm::relay::WildcardPattern</a></td><td class="desc">A pattern which matches anything </td></tr>
-<tr id="row_113_195_" 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="classtvm_1_1relay_1_1DFPatternCallback.html" target="_self">tvm::relay::DFPatternCallback</a></td><td class="desc">Managed reference to dataflow pattern callbacks </td></tr>
-<tr id="row_113_196_" 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="classtvm_1_1relay_1_1Executor.html" target="_self">tvm::relay::Executor</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1relay_1_1ExecutorNode.html" title="Executor information. ">ExecutorNode</a> </td></tr>
-<tr id="row_113_197_" 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="classtvm_1_1relay_1_1Id.html" target="_self">tvm::relay::Id</a></td><td class="desc"></td></tr>
-<tr id="row_113_198_" 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="classtvm_1_1relay_1_1OpImplementation.html" target="_self">tvm::relay::OpImplementation</a></td><td class="desc">Operator implementation class </td></tr>
-<tr id="row_113_199_" 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="classtvm_1_1relay_1_1OpSpecialization.html" target="_self">tvm::relay::OpSpecialization</a></td><td class="desc">Operator specialization class </td></tr>
-<tr id="row_113_200_" 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="classtvm_1_1relay_1_1OpStrategy.html" target="_self">tvm::relay::OpStrategy</a></td><td class="desc">Operator strategy class </td></tr>
-<tr id="row_113_201_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_201_" class="arrow" onclick="toggleFolder('113_201_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1Pattern.html" target="_self">tvm::relay::Pattern</a></td><td class="desc"><a class="el" href="classtvm_1_1relay_1_1Pattern.html" title="Pattern is the base type for an ADT mat [...]
-<tr id="row_113_201_0_" 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="classtvm_1_1relay_1_1PatternConstructor.html" target="_self">tvm::relay::PatternConstructor</a></td><td class="desc"></td></tr>
-<tr id="row_113_201_1_" 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="classtvm_1_1relay_1_1PatternTuple.html" target="_self">tvm::relay::PatternTuple</a></td><td class="desc"></td></tr>
-<tr id="row_113_201_2_" 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="classtvm_1_1relay_1_1PatternVar.html" target="_self">tvm::relay::PatternVar</a></td><td class="desc"></td></tr>
-<tr id="row_113_201_3_" 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="classtvm_1_1relay_1_1PatternWildcard.html" target="_self">tvm::relay::PatternWildcard</a></td><td class="desc"></td></tr>
-<tr id="row_113_202_" 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="classtvm_1_1relay_1_1RecClosure.html" target="_self">tvm::relay::RecClosure</a></td><td class="desc"></td></tr>
-<tr id="row_113_203_" 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="classtvm_1_1relay_1_1RefValue.html" target="_self">tvm::relay::RefValue</a></td><td class="desc"></td></tr>
-<tr id="row_113_204_" 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="classtvm_1_1relay_1_1Runtime.html" target="_self">tvm::relay::Runtime</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1relay_1_1RuntimeNode.html" title="Runtime information. ">RuntimeNode</a> </td></tr>
-<tr id="row_113_205_" 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="classtvm_1_1runtime_1_1ADT.html" target="_self">tvm::runtime::ADT</a></td><td class="desc">Reference to algebraic data type objects </td></tr>
-<tr id="row_113_206_" 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="classtvm_1_1runtime_1_1Array.html" target="_self">tvm::runtime::Array< T, typename ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a>, container representing a [...]
-<tr id="row_113_207_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_207_" class="arrow" onclick="toggleFolder('113_207_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1Closure.html" target="_self">tvm::runtime::Closure</a></td><td class="desc">Reference to closure </td></tr>
-<tr id="row_113_207_0_" 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="classtvm_1_1relay_1_1InterpreterClosure.html" target="_self">tvm::relay::InterpreterClosure</a></td><td class="desc"></td></tr>
-<tr id="row_113_207_1_" 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="classtvm_1_1runtime_1_1vm_1_1VMClosure.html" target="_self">tvm::runtime::vm::VMClosure</a></td><td class="desc">Reference to closure </td></tr>
-<tr id="row_113_208_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< K, V, typename, typename ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Map.html" title="Map container of NodeRef->NodeRef in DSL graph. Map implements copy on write semantics, which m [...]
-<tr id="row_113_209_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_209_" class="arrow" onclick="toggleFolder('113_209_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html" target="_self">tvm::runtime::metadata::MetadataBase</a></td><td class="desc">Reference class for the common <a class="el" href="classtvm_1_1ru [...]
-<tr id="row_113_209_0_" 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="classtvm_1_1runtime_1_1metadata_1_1ConstantInfoMetadata.html" target="_self">tvm::runtime::metadata::ConstantInfoMetadata</a></td><td class="desc"></td></tr>
-<tr id="row_113_209_1_" 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="classtvm_1_1runtime_1_1metadata_1_1Metadata.html" target="_self">tvm::runtime::metadata::Metadata</a></td><td class="desc"></td></tr>
-<tr id="row_113_209_2_" 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="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" target="_self">tvm::runtime::metadata::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. ">MetadataA [...]
-<tr id="row_113_209_3_" 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="classtvm_1_1runtime_1_1metadata_1_1TensorInfo.html" target="_self">tvm::runtime::metadata::TensorInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_210_" 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="classtvm_1_1runtime_1_1Module.html" target="_self">tvm::runtime::Module</a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Module.html" title="Module container of TVM. ">Module</a> container of TVM </td></tr>
-<tr id="row_113_211_" 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="classtvm_1_1runtime_1_1NDArray.html" target="_self">tvm::runtime::NDArray</a></td><td class="desc">Managed <a class="el" href="classtvm_1_1runtime_1_1NDArray.html" title="Managed NDArray. The array is backed by reference counted blocks. ">NDArray</a>. The array is backed by referenc [...]
-<tr id="row_113_212_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< T ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Optional.html" title="Optional container that to represent to a Nullable variant of T. ">Optional</a> container that to represen [...]
-<tr id="row_113_213_" 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="classtvm_1_1runtime_1_1PackedFunc.html" target="_self">tvm::runtime::PackedFunc</a></td><td class="desc">Packed function is a type-erased function. The arguments are passed by packed format </td></tr>
-<tr id="row_113_214_" 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="classtvm_1_1runtime_1_1profiling_1_1DeviceWrapper.html" target="_self">tvm::runtime::profiling::DeviceWrapper</a></td><td class="desc">Wrapper for <code>Device</code> </td></tr>
-<tr id="row_113_215_" 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="classtvm_1_1runtime_1_1profiling_1_1MetricCollector.html" target="_self">tvm::runtime::profiling::MetricCollector</a></td><td class="desc">Wrapper for <code><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollectorNode.html" title="Interface for user defined profiling [...]
-<tr id="row_113_216_" 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="classtvm_1_1runtime_1_1profiling_1_1Report.html" target="_self">tvm::runtime::profiling::Report</a></td><td class="desc"></td></tr>
-<tr id="row_113_217_" 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="classtvm_1_1runtime_1_1ShapeTuple.html" target="_self">tvm::runtime::ShapeTuple</a></td><td class="desc">Reference to shape tuple objects </td></tr>
-<tr id="row_113_218_" 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="classtvm_1_1runtime_1_1String.html" target="_self">tvm::runtime::String</a></td><td class="desc">Reference to string objects </td></tr>
-<tr id="row_113_219_" 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="classtvm_1_1runtime_1_1Timer.html" target="_self">tvm::runtime::Timer</a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Timer.html" title="Timer for a specific device. ">Timer</a> for a specific device </td></tr>
-<tr id="row_113_220_" 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="classtvm_1_1runtime_1_1vm_1_1Storage.html" target="_self">tvm::runtime::vm::Storage</a></td><td class="desc">Reference to storage </td></tr>
-<tr id="row_113_221_" 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="classtvm_1_1script_1_1ir__builder_1_1IRBuilder.html" target="_self">tvm::script::ir_builder::IRBuilder</a></td><td class="desc">Managed reference to an <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderNode.html" title="A dialect-agnostic IRBuilder that constructs an [...]
-<tr id="row_113_222_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_222_" class="arrow" onclick="toggleFolder('113_222_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderFrame.html" target="_self">tvm::script::ir_builder::IRBuilderFrame</a></td><td class="desc">Managed reference to an <a class="el" href="classtvm_1_1sc [...]
-<tr id="row_113_222_0_" 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="classtvm_1_1script_1_1ir__builder_1_1ir_1_1IRModuleFrame.html" target="_self">tvm::script::ir_builder::ir::IRModuleFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1ir_1_1IRModuleFrameNode.html" title="A frame that represe [...]
-<tr id="row_113_222_1_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_222_1_" class="arrow" onclick="toggleFolder('113_222_1_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1TIRFrame.html" target="_self">tvm::script::ir_builder::tir::TIRFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_ [...]
-<tr id="row_113_222_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_1script_1_1ir__builder_1_1tir_1_1AllocateConstFrame.html" target="_self">tvm::script::ir_builder::tir::AllocateConstFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AllocateConstFrameNode.html" title=" [...]
-<tr id="row_113_222_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_1script_1_1ir__builder_1_1tir_1_1AllocateFrame.html" target="_self">tvm::script::ir_builder::tir::AllocateFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AllocateFrameNode.html" title="A frame represe [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1AssertFrame.html" target="_self">tvm::script::ir_builder::tir::AssertFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AssertFrameNode.html" title="A frame that represen [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1AttrFrame.html" target="_self">tvm::script::ir_builder::tir::AttrFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AttrFrameNode.html" title="A frame that represents att [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1BlockFrame.html" target="_self">tvm::script::ir_builder::tir::BlockFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1BlockFrameNode.html" title="A frame that represents [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1BlockInitFrame.html" target="_self">tvm::script::ir_builder::tir::BlockInitFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1BlockInitFrameNode.html" title="A frame that [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1DeclBufferFrame.html" target="_self">tvm::script::ir_builder::tir::DeclBufferFrame</a></td><td class="desc"></td></tr>
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1ElseFrame.html" target="_self">tvm::script::ir_builder::tir::ElseFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrameNode.html" title="A frame that represents els [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1ForFrame.html" target="_self">tvm::script::ir_builder::tir::ForFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ForFrameNode.html" title="A frame that represents the fo [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1IfFrame.html" target="_self">tvm::script::ir_builder::tir::IfFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1IfFrameNode.html" title="A frame that represents if statem [...]
-<tr id="row_113_222_1_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_1script_1_1ir__builder_1_1tir_1_1LaunchThreadFrame.html" target="_self">tvm::script::ir_builder::tir::LaunchThreadFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1LaunchThreadFrameNode.html" title="Th [...]
-<tr id="row_113_222_1_11_" 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_1script_1_1ir__builder_1_1tir_1_1LetFrame.html" target="_self">tvm::script::ir_builder::tir::LetFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1LetFrameNode.html" title="A frame represents the let bi [...]
-<tr id="row_113_222_1_12_" 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_1script_1_1ir__builder_1_1tir_1_1PrimFuncFrame.html" target="_self">tvm::script::ir_builder::tir::PrimFuncFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1PrimFuncFrameNode.html" title="A frame that r [...]
-<tr id="row_113_222_1_13_" 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_1script_1_1ir__builder_1_1tir_1_1RealizeFrame.html" target="_self">tvm::script::ir_builder::tir::RealizeFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1RealizeFrameNode.html" title="A frame that repr [...]
-<tr id="row_113_222_1_14_" 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_1script_1_1ir__builder_1_1tir_1_1ThenFrame.html" target="_self">tvm::script::ir_builder::tir::ThenFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ThenFrameNode.html" title="A frame that represents th [...]
-<tr id="row_113_222_1_15_" 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_1script_1_1ir__builder_1_1tir_1_1WhileFrame.html" target="_self">tvm::script::ir_builder::tir::WhileFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1WhileFrameNode.html" title="A frame that represents [...]
-<tr id="row_113_223_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_223_" class="arrow" onclick="toggleFolder('113_223_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1Doc.html" target="_self">tvm::script::printer::Doc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1DocNode.html" ti [...]
-<tr id="row_113_223_0_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_223_0_" class="arrow" onclick="toggleFolder('113_223_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1ExprDoc.html" target="_self">tvm::script::printer::ExprDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1Ex [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1AttrAccessDoc.html" target="_self">tvm::script::printer::AttrAccessDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AttrAccessDocNode.html" title="Doc that represents attribute access on another [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1CallDoc.html" target="_self">tvm::script::printer::CallDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1CallDocNode.html" title="Doc that represents function call. ">CallDocNode</a> </td></tr>
-<tr id="row_113_223_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_1script_1_1printer_1_1DictDoc.html" target="_self">tvm::script::printer::DictDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1DictDocNode.html" title="Doc that represents dictionary literal. ">DictDocNode</a> </td></tr>
-<tr id="row_113_223_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_1script_1_1printer_1_1IdDoc.html" target="_self">tvm::script::printer::IdDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IdDocNode.html" title="Doc that represents identifier. ">IdDocNode</a> </td></tr>
-<tr id="row_113_223_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_1script_1_1printer_1_1IndexDoc.html" target="_self">tvm::script::printer::IndexDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IndexDocNode.html" title="Doc that represents index access on another expression. ">Index [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1LambdaDoc.html" target="_self">tvm::script::printer::LambdaDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1LambdaDocNode.html" title="Doc that represents anonymous function. ">LambdaDocNode</a> [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1ListDoc.html" target="_self">tvm::script::printer::ListDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ListDocNode.html" title="Doc that represents list literal. ">ListDocNode</a> </td></tr>
-<tr id="row_113_223_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_1script_1_1printer_1_1LiteralDoc.html" target="_self">tvm::script::printer::LiteralDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1LiteralDocNode.html" title="Doc that represents literal value. ">LiteralDocNode</a> < [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1OperationDoc.html" target="_self">tvm::script::printer::OperationDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1OperationDocNode.html" title="Doc that represents operation. ">OperationDocNode</ [...]
-<tr id="row_113_223_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_1script_1_1printer_1_1TupleDoc.html" target="_self">tvm::script::printer::TupleDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1TupleDocNode.html" title="Doc that represents tuple literal. ">TupleDocNode</a> </td></tr>
-<tr id="row_113_223_1_" 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="classtvm_1_1script_1_1printer_1_1SliceDoc.html" target="_self">tvm::script::printer::SliceDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1SliceDocNode.html" title="Doc that represents slice in Index expression. ">SliceDocNode</a> [...]
-<tr id="row_113_223_2_" 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="classtvm_1_1script_1_1printer_1_1StmtBlockDoc.html" target="_self">tvm::script::printer::StmtBlockDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1StmtBlockDocNode.html" title="The container doc that holds a list of StmtDoc. ">Stm [...]
-<tr id="row_113_223_3_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_223_3_" class="arrow" onclick="toggleFolder('113_223_3_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1StmtDoc.html" target="_self">tvm::script::printer::StmtDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1St [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1AssertDoc.html" target="_self">tvm::script::printer::AssertDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AssertDocNode.html" title="Doc that represents assert statement. ">AssertDocNode</a> </ [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1AssignDoc.html" target="_self">tvm::script::printer::AssignDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AssignDocNode.html" title="Doc that represents assign statement. ">AssignDocNode</a> </ [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1ClassDoc.html" target="_self">tvm::script::printer::ClassDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ClassDocNode.html" title="Doc that represents class definition. ">ClassDocNode</a> </td></tr>
-<tr id="row_113_223_3_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_1script_1_1printer_1_1ExprStmtDoc.html" target="_self">tvm::script::printer::ExprStmtDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ExprStmtDocNode.html" title="Doc that represents an expression as statement. ">Expr [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1ForDoc.html" target="_self">tvm::script::printer::ForDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ForDocNode.html" title="Doc that represents for statement. ">ForDocNode</a> </td></tr>
-<tr id="row_113_223_3_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_1script_1_1printer_1_1FunctionDoc.html" target="_self">tvm::script::printer::FunctionDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1FunctionDocNode.html" title="Doc that represents function definition. ">FunctionDoc [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1IfDoc.html" target="_self">tvm::script::printer::IfDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IfDocNode.html" title="Doc that represent if-then-else statement. ">IfDocNode</a> </td></tr>
-<tr id="row_113_223_3_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_1script_1_1printer_1_1ReturnDoc.html" target="_self">tvm::script::printer::ReturnDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ReturnDocNode.html" title="Doc that represents return statement. ">ReturnDocNode</a> </ [...]
-<tr id="row_113_223_3_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_1script_1_1printer_1_1ScopeDoc.html" target="_self">tvm::script::printer::ScopeDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ScopeDocNode.html" title="Doc that represents special scopes. ">ScopeDocNode</a> </td></tr>
-<tr id="row_113_223_3_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_1script_1_1printer_1_1WhileDoc.html" target="_self">tvm::script::printer::WhileDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1WhileDocNode.html" title="Doc that represents while statement. ">WhileDocNode</a> </td></tr>
-<tr id="row_113_224_" 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="classtvm_1_1script_1_1printer_1_1Frame.html" target="_self">tvm::script::printer::Frame</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1FrameNode.html">FrameNode</a> </td></tr>
-<tr id="row_113_225_" 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="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" target="_self">tvm::script::printer::IRDocsifier</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html" title="IRDocsifier is the top-level interface in the IR->Doc p [...]
-<tr id="row_113_226_" 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="classtvm_1_1SourceName.html" target="_self">tvm::SourceName</a></td><td class="desc">The source name of a file span </td></tr>
-<tr id="row_113_227_" 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="classtvm_1_1Span.html" target="_self">tvm::Span</a></td><td class="desc"></td></tr>
-<tr id="row_113_228_" 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="classtvm_1_1Target.html" target="_self">tvm::Target</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetNode.html" title="Compilation target. ">TargetNode</a> </td></tr>
-<tr id="row_113_229_" 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="classtvm_1_1TargetKind.html" target="_self">tvm::TargetKind</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetKindNode.html" title="Target kind, specifies the kind of the target. ">TargetKindNode</a> </td></tr>
-<tr id="row_113_230_" 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="classtvm_1_1TargetTag.html" target="_self">tvm::TargetTag</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetTagNode.html" title="A target tag. ">TargetTagNode</a> </td></tr>
-<tr id="row_113_231_" 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="classtvm_1_1te_1_1IterVarAttr.html" target="_self">tvm::te::IterVarAttr</a></td><td class="desc">Additional scheduable attributes about IterVar </td></tr>
-<tr id="row_113_232_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_232_" class="arrow" onclick="toggleFolder('113_232_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1te_1_1IterVarRelation.html" target="_self">tvm::te::IterVarRelation</a></td><td class="desc">The schedule relation between IterVars can be <a class="el" href="classtvm_1_1te_1_1Split.h [...]
-<tr id="row_113_232_0_" 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="classtvm_1_1te_1_1Fuse.html" target="_self">tvm::te::Fuse</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1FuseNode.html" title="Fuse two domains into one domain. ">FuseNode</a> </td></tr>
-<tr id="row_113_232_1_" 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="classtvm_1_1te_1_1Rebase.html" target="_self">tvm::te::Rebase</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1RebaseNode.html" title="Rebase the iteration to make min to be 0. This is useful to normalize the Schedule to make every leaf...">Rebas [...]
-<tr id="row_113_232_2_" 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="classtvm_1_1te_1_1Singleton.html" target="_self">tvm::te::Singleton</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1SingletonNode.html" title="Singleton iterator [0, 1) ">SingletonNode</a> </td></tr>
-<tr id="row_113_232_3_" 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="classtvm_1_1te_1_1Split.html" target="_self">tvm::te::Split</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1SplitNode.html" title="Split the parent domain into product of outer and iter. ">SplitNode</a> </td></tr>
-<tr id="row_113_232_4_" 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="classtvm_1_1te_1_1Transform.html" target="_self">tvm::te::Transform</a></td><td class="desc"></td></tr>
-<tr id="row_113_233_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_233_" class="arrow" onclick="toggleFolder('113_233_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1te_1_1Operation.html" target="_self">tvm::te::Operation</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Operation.html" title="Operation that produces tensors. ">Operati [...]
-<tr id="row_113_233_0_" 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="classtvm_1_1te_1_1ComputeOp.html" target="_self">tvm::te::ComputeOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ComputeOpNode.html" title="A Compute op that compute a tensor on certain domain. ">ComputeOpNode</a> </td></tr>
-<tr id="row_113_233_1_" 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="classtvm_1_1te_1_1ExternOp.html" target="_self">tvm::te::ExternOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ExternOpNode.html" title="External computation that cannot be splitted. ">ExternOpNode</a> </td></tr>
-<tr id="row_113_233_2_" 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="classtvm_1_1te_1_1HybridOp.html" target="_self">tvm::te::HybridOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1HybridOpNode.html" title="A computation operator that generated by hybrid script. ">HybridOpNode</a> </td></tr>
-<tr id="row_113_233_3_" 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="classtvm_1_1te_1_1PlaceholderOp.html" target="_self">tvm::te::PlaceholderOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1PlaceholderOpNode.html" title="A placeholder op represents an input placeholder. ">PlaceholderOpNode</a> </td></tr>
-<tr id="row_113_233_4_" 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="classtvm_1_1te_1_1ScanOp.html" target="_self">tvm::te::ScanOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ScanOpNode.html" title="Symbolic scan. ">ScanOpNode</a> </td></tr>
-<tr id="row_113_233_5_" 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="classtvm_1_1te_1_1TensorComputeOp.html" target="_self">tvm::te::TensorComputeOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorComputeOpNode.html" title="A TenorCompute op that compute a tensor with an tensor intrinsic. ">TensorComputeOpN [...]
-<tr id="row_113_234_" 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="classtvm_1_1te_1_1Schedule.html" target="_self">tvm::te::Schedule</a></td><td class="desc">Global schedule container For operations and all the operations they depend on. The schedule per <a class="el" href="classtvm_1_1te_1_1Operation.html" title="Operation that produces tensors. " [...]
-<tr id="row_113_235_" 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="classtvm_1_1te_1_1SpecializedCondition.html" target="_self">tvm::te::SpecializedCondition</a></td><td class="desc">Specialized condition to enable op specialization </td></tr>
-<tr id="row_113_236_" 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="classtvm_1_1te_1_1Stage.html" target="_self">tvm::te::Stage</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Stage.html" title="Stage, contains scheduling for a stage of computation. ">Stage</a>, contains scheduling for a stage of computation </td></tr>
-<tr id="row_113_237_" 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="classtvm_1_1te_1_1TensorIntrin.html" target="_self">tvm::te::TensorIntrin</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorIntrinNode.html" title="Node to represent a Tensor intrinsic operator. ">TensorIntrinNode</a> </td></tr>
-<tr id="row_113_238_" 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="classtvm_1_1te_1_1TensorIntrinCall.html" target="_self">tvm::te::TensorIntrinCall</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorIntrinCallNode.html">TensorIntrinCallNode</a> </td></tr>
-<tr id="row_113_239_" 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="classtvm_1_1tir_1_1BijectiveLayout.html" target="_self">tvm::tir::BijectiveLayout</a></td><td class="desc">Bijective function mapping for data layout transformation. Given two <a class="el" href="classtvm_1_1tir_1_1Layout.html" title="Managed reference to LayoutNode. ">Layout</a>, < [...]
-<tr id="row_113_240_" 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="classtvm_1_1tir_1_1BlockRV.html" target="_self">tvm::tir::BlockRV</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockRVNode.html" title="A random variable that evaluates to a TensorIR block. ">BlockRVNode</a> </td></tr>
-<tr id="row_113_241_" 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="classtvm_1_1tir_1_1BlockScope.html" target="_self">tvm::tir::BlockScope</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockScopeNode.html" title="An object with 1-to-1 correspondence with each block reference in the sref tree. This data structu [...]
-<tr id="row_113_242_" 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="classtvm_1_1tir_1_1Buffer.html" target="_self">tvm::tir::Buffer</a></td><td class="desc"><a class="el" href="classtvm_1_1tir_1_1Buffer.html" title="Buffer is a symbolic n-darray structure. It is a composition of primitive symbolic types...">Buffer</a> is a symbolic n-darray structur [...]
-<tr id="row_113_243_" 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="classtvm_1_1tir_1_1BufferRegion.html" target="_self">tvm::tir::BufferRegion</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferRegionNode.html" title="Representing the region of multi-dimensional buffer access. ">BufferRegionNode</a> </td></tr>
-<tr id="row_113_244_" 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="classtvm_1_1tir_1_1CommReducer.html" target="_self">tvm::tir::CommReducer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CommReducerNode.html" title="A commutative reducer node to represent a commutative binary operator with identity element..." [...]
-<tr id="row_113_245_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_245_" class="arrow" onclick="toggleFolder('113_245_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1DataProducer.html" target="_self">tvm::tir::DataProducer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DataProducerNode.html" title="Base [...]
-<tr id="row_113_245_0_" 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="classtvm_1_1te_1_1Tensor.html" target="_self">tvm::te::Tensor</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Tensor.html" title="Tensor structure representing a possible input, or intermediate computation result. ">Tensor</a> structure representing a possible input [...]
-<tr id="row_113_246_" 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="classtvm_1_1tir_1_1Dependency.html" target="_self">tvm::tir::Dependency</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DependencyNode.html" title="A tuple (src, dst, kind) representing certain types of dependency. For example, (A, B, kRAW) means [...]
-<tr id="row_113_247_" 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="classtvm_1_1tir_1_1IndexMap.html" target="_self">tvm::tir::IndexMap</a></td><td class="desc"></td></tr>
-<tr id="row_113_248_" 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="classtvm_1_1tir_1_1Instruction.html" target="_self">tvm::tir::Instruction</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1InstructionNode.html" title="Schedule instructions each corresponds to a schedule primitive. ">InstructionNode</a> </td></tr>
-<tr id="row_113_249_" 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="classtvm_1_1tir_1_1InstructionKind.html" target="_self">tvm::tir::InstructionKind</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1InstructionKindNode.html" title="Kind of an instruction, e.g. Split, Reorder, etc. Besides the name, every kind of i [...]
-<tr id="row_113_250_" 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="classtvm_1_1tir_1_1IterVar.html" target="_self">tvm::tir::IterVar</a></td><td class="desc">Iteration Variable, represents an iteration over an integer interval </td></tr>
-<tr id="row_113_251_" 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="classtvm_1_1tir_1_1Layout.html" target="_self">tvm::tir::Layout</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LayoutNode.html" title="Layout is to describe how data is organized within an N-dimention tensor. It is composed of upper cas...">Layo [...]
-<tr id="row_113_252_" 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="classtvm_1_1tir_1_1LoopRV.html" target="_self">tvm::tir::LoopRV</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LoopRVNode.html" title="A random variable that evaluates to a TensorIR for loop. ">LoopRVNode</a> </td></tr>
-<tr id="row_113_253_" 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="classtvm_1_1tir_1_1MatchBufferRegion.html" target="_self">tvm::tir::MatchBufferRegion</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MatchBufferRegionNode.html" title="Match introduces a constraint that the source buffer region can be remapped t [...]
-<tr id="row_113_254_" 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="classtvm_1_1tir_1_1Schedule.html" target="_self">tvm::tir::Schedule</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ScheduleNode.html" title="The user-facing schedule class. ">ScheduleNode</a> </td></tr>
-<tr id="row_113_255_" 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="classtvm_1_1tir_1_1ScheduleState.html" target="_self">tvm::tir::ScheduleState</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ScheduleStateNode.html" title="The state of scheduling, which exposes a Replace method as the primary interface for all [...]
-<tr id="row_113_256_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_256_" class="arrow" onclick="toggleFolder('113_256_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1Stmt.html" target="_self">tvm::tir::Stmt</a></td><td class="desc">Container of all statements </td></tr>
-<tr id="row_113_256_0_" 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="classtvm_1_1tir_1_1Allocate.html" target="_self">tvm::tir::Allocate</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AllocateNode.html" title="Allocate a buffer that can be used in body. ">AllocateNode</a> </td></tr>
-<tr id="row_113_256_1_" 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="classtvm_1_1tir_1_1AllocateConst.html" target="_self">tvm::tir::AllocateConst</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AllocateConstNode.html" title="Allocate a buffer that can be used in body. ">AllocateConstNode</a> </td></tr>
-<tr id="row_113_256_2_" 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="classtvm_1_1tir_1_1AssertStmt.html" target="_self">tvm::tir::AssertStmt</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AssertStmtNode.html" title="Assert condition, if an error occurs, return the error message. ">AssertStmtNode</a> </td></tr>
-<tr id="row_113_256_3_" 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="classtvm_1_1tir_1_1AttrStmt.html" target="_self">tvm::tir::AttrStmt</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AttrStmtNode.html" title="Define certain auxiliary attribute for the body to be a symbolic value. This provide auxiliary inform. [...]
-<tr id="row_113_256_4_" 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="classtvm_1_1tir_1_1Block.html" target="_self">tvm::tir::Block</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockNode.html" title="A block is a basic schedule unit in TIR. ">BlockNode</a> </td></tr>
-<tr id="row_113_256_5_" 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="classtvm_1_1tir_1_1BlockRealize.html" target="_self">tvm::tir::BlockRealize</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockRealizeNode.html" title="A block realization node represents execution of the block at the binding values. ...">Blo [...]
-<tr id="row_113_256_6_" 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="classtvm_1_1tir_1_1BufferRealize.html" target="_self">tvm::tir::BufferRealize</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferRealizeNode.html" title="Annotate the region where the buffer need to be read and write in the body. We only nee [...]
-<tr id="row_113_256_7_" 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="classtvm_1_1tir_1_1BufferStore.html" target="_self">tvm::tir::BufferStore</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferStoreNode.html" title="Store value to the high dimension buffer. ">BufferStoreNode</a> </td></tr>
-<tr id="row_113_256_8_" 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="classtvm_1_1tir_1_1DeclBuffer.html" target="_self">tvm::tir::DeclBuffer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DeclBufferNode.html" title="Declare a buffer that can be used in the body. ">DeclBufferNode</a> </td></tr>
-<tr id="row_113_256_9_" 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="classtvm_1_1tir_1_1Evaluate.html" target="_self">tvm::tir::Evaluate</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1EvaluateNode.html" title="Evaluates an expression. This is mostly used for putting a Call node into Stmt. ">EvaluateNode</a> </td></tr>
-<tr id="row_113_256_10_" 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="classtvm_1_1tir_1_1For.html" target="_self">tvm::tir::For</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ForNode.html" title="A for loop, with poissible type annotations. ">ForNode</a> </td></tr>
-<tr id="row_113_256_11_" 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="classtvm_1_1tir_1_1IfThenElse.html" target="_self">tvm::tir::IfThenElse</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1IfThenElseNode.html" title="IfThenElse statment. ">IfThenElseNode</a> </td></tr>
-<tr id="row_113_256_12_" 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="classtvm_1_1tir_1_1LetStmt.html" target="_self">tvm::tir::LetStmt</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LetStmtNode.html" title="Let binding, bind var to value, then run body. ">LetStmtNode</a> </td></tr>
-<tr id="row_113_256_13_" 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="classtvm_1_1tir_1_1Prefetch.html" target="_self">tvm::tir::Prefetch</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1PrefetchNode.html" title="A prefetch hint for a buffer. ">PrefetchNode</a> </td></tr>
-<tr id="row_113_256_14_" 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="classtvm_1_1tir_1_1ProducerRealize.html" target="_self">tvm::tir::ProducerRealize</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ProducerRealizeNode.html" title="Annotate the bounds where the data produced by the producer need to be written a [...]
-<tr id="row_113_256_15_" 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="classtvm_1_1tir_1_1ProducerStore.html" target="_self">tvm::tir::ProducerStore</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ProducerStoreNode.html" title="Store value into mult-dimensional array that will be read by the consumer of the produ [...]
-<tr id="row_113_256_16_" 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="classtvm_1_1tir_1_1SeqStmt.html" target="_self">tvm::tir::SeqStmt</a></td><td class="desc">Sequence statement </td></tr>
-<tr id="row_113_256_17_" 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="classtvm_1_1tir_1_1Store.html" target="_self">tvm::tir::Store</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1StoreNode.html" title="Store value to the buffer. ">StoreNode</a> </td></tr>
-<tr id="row_113_256_18_" 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="classtvm_1_1tir_1_1While.html" target="_self">tvm::tir::While</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1WhileNode.html" title="A While loop. ">WhileNode</a> </td></tr>
-<tr id="row_113_257_" 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="classtvm_1_1tir_1_1StmtSRef.html" target="_self">tvm::tir::StmtSRef</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1StmtSRefNode.html" title="An object that refers to schedulable elements (block/for-loop) in TensorIR, aka "sref". ">Stmt [...]
-<tr id="row_113_258_" 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="classtvm_1_1tir_1_1TensorIntrin.html" target="_self">tvm::tir::TensorIntrin</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1TensorIntrinNode.html" title="Tensor intrinsics for tensorization. ">TensorIntrinNode</a> </td></tr>
-<tr id="row_113_259_" 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="classtvm_1_1tir_1_1Trace.html" target="_self">tvm::tir::Trace</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1TraceNode.html" title="An execution trace of a scheduling program. ">TraceNode</a> </td></tr>
-<tr id="row_113_260_" 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="classtvm_1_1tir_1_1usmp_1_1AllocatedPoolInfo.html" target="_self">tvm::tir::usmp::AllocatedPoolInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_261_" 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="classtvm_1_1tir_1_1usmp_1_1BufferInfo.html" target="_self">tvm::tir::usmp::BufferInfo</a></td><td class="desc"></td></tr>
-<tr id="row_113_262_" 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="classtvm_1_1tir_1_1usmp_1_1BufferInfoAnalysis.html" target="_self">tvm::tir::usmp::BufferInfoAnalysis</a></td><td class="desc"></td></tr>
-<tr id="row_113_263_" 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="classtvm_1_1tir_1_1usmp_1_1PoolAllocation.html" target="_self">tvm::tir::usmp::PoolAllocation</a></td><td class="desc"></td></tr>
-<tr id="row_113_264_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_264_" class="arrow" onclick="toggleFolder('113_264_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1transform_1_1Pass.html" target="_self">tvm::transform::Pass</a></td><td class="desc"></td></tr>
-<tr id="row_113_264_0_" 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="classtvm_1_1transform_1_1Sequential.html" target="_self">tvm::transform::Sequential</a></td><td class="desc"></td></tr>
-<tr id="row_113_265_" 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="classtvm_1_1transform_1_1PassContext.html" target="_self">tvm::transform::PassContext</a></td><td class="desc"><a class="el" href="classtvm_1_1transform_1_1PassContext.html" title="PassContext that is used to configure the pass behavior. ">PassContext</a> that is used to configure t [...]
-<tr id="row_113_266_" 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="classtvm_1_1transform_1_1PassInfo.html" target="_self">tvm::transform::PassInfo</a></td><td class="desc">Managed reference class for <a class="el" href="classtvm_1_1transform_1_1PassInfoNode.html" title="Meta data that will be used to help optimization and analysis. ">PassInfoNode</ [...]
-<tr id="row_113_267_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_267_" class="arrow" onclick="toggleFolder('113_267_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Type.html" target="_self">tvm::Type</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TypeNode.html" title="Type is the base type of all types. ">TypeNode</ [...]
-<tr id="row_113_267_0_" 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="classtvm_1_1BaseTensorType.html" target="_self">tvm::BaseTensorType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseTensorTypeNode.html" title="Base of all Tensor types This container can hold TensorType or GenericTensorType. ...">BaseTensorTypeNo [...]
-<tr id="row_113_267_1_" 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="classtvm_1_1FuncType.html" target="_self">tvm::FuncType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1FuncTypeNode.html" title="Function type. ">FuncTypeNode</a> </td></tr>
-<tr id="row_113_267_2_" 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="classtvm_1_1GlobalTypeVar.html" target="_self">tvm::GlobalTypeVar</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1GlobalTypeVarNode.html" title="A global type variable that is used for defining new types or type aliases. ">GlobalTypeVarNode</a> </td></tr>
-<tr id="row_113_267_3_" 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="classtvm_1_1IncompleteType.html" target="_self">tvm::IncompleteType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1IncompleteTypeNode.html" title="Intermediate values that is used to indicate incomplete type during type inference. ">IncompleteTypeNod [...]
-<tr id="row_113_267_4_" 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="classtvm_1_1PointerType.html" target="_self">tvm::PointerType</a></td><td class="desc"></td></tr>
-<tr id="row_113_267_5_" 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="classtvm_1_1PrimType.html" target="_self">tvm::PrimType</a></td><td class="desc"></td></tr>
-<tr id="row_113_267_6_" 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="classtvm_1_1RelayRefType.html" target="_self">tvm::RelayRefType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1RelayRefTypeNode.html" title="Reference Type High-level Relay IR. ">RelayRefTypeNode</a> </td></tr>
-<tr id="row_113_267_7_" 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="classtvm_1_1TensorType.html" target="_self">tvm::TensorType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TensorTypeNode.html" title="This is the most commonly used type in relay. TensorType have a fixed dimension, data type...">TensorTypeNode</a> < [...]
-<tr id="row_113_267_8_" 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="classtvm_1_1TupleType.html" target="_self">tvm::TupleType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TupleTypeNode.html" title="The type of tuple values. ">TupleTypeNode</a> </td></tr>
-<tr id="row_113_267_9_" 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="classtvm_1_1TypeCall.html" target="_self">tvm::TypeCall</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TypeCallNode.html" title="Type function application. ">TypeCallNode</a> </td></tr>
-<tr id="row_113_267_10_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_267_10_" class="arrow" onclick="toggleFolder('113_267_10_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1TypeConstraint.html" target="_self">tvm::TypeConstraint</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TypeConstraintNode.html" title="Potential [...]
-<tr id="row_113_267_10_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_1TypeRelation.html" target="_self">tvm::TypeRelation</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TypeRelationNode.html" title="User defined type relation, it is an input-output relation on types. ">TypeRelationNode</a> </td></tr>
-<tr id="row_113_267_11_" 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="classtvm_1_1TypeData.html" target="_self">tvm::TypeData</a></td><td class="desc">Stores all data for an Algebraic Data <a class="el" href="classtvm_1_1Type.html" title="Managed reference to TypeNode. ">Type</a> (ADT) </td></tr>
-<tr id="row_113_267_12_" 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="classtvm_1_1TypeVar.html" target="_self">tvm::TypeVar</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1TypeVarNode.html" title="Type parameter in functions. ">TypeVarNode</a> </td></tr>
-<tr id="row_113_268_" 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="classtvm_1_1TypedEnvFunc_3_01R_07Args_8_8_8_08_4.html" target="_self">tvm::TypedEnvFunc< R(Args...)></a></td><td class="desc">A typed version of <a class="el" href="classtvm_1_1EnvFunc.html" title="Managed reference to EnvFuncNode. ">EnvFunc</a>. It is backed by a GlobalFuncNo [...]
-<tr id="row_113_269_" 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="classtvm_1_1TypeReporter.html" target="_self">tvm::TypeReporter</a></td><td class="desc">Container class of <a class="el" href="classtvm_1_1TypeReporter.html" title="Container class of TypeReporter. ">TypeReporter</a> </td></tr>
-<tr id="row_113_270_" 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="classtvm_1_1VirtualDevice.html" target="_self">tvm::VirtualDevice</a></td><td class="desc">Managed reference class to <code><a class="el" href="classtvm_1_1VirtualDeviceNode.html" title="Describes at compile time the constraints on where data is to be stored at runtime down to the ( [...]
-<tr id="row_113_271_" 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="classtvm_1_1WorkspaceMemoryPools.html" target="_self">tvm::WorkspaceMemoryPools</a></td><td class="desc"></td></tr>
+<tr id="row_113_71_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::SourceName, tvm::parser::Source ></a></td><td class="desc"></td></tr>
+<tr id="row_113_72_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::Target, tvm::Integer ></a></td><td class="desc"></td></tr>
+<tr id="row_113_73_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::te::Operation, tvm::te::Stage ></a></td><td class="desc"></td></tr>
+<tr id="row_113_74_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Buffer, tvm::tir::Buffer ></a></td><td class="desc"></td></tr>
+<tr id="row_113_75_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Instruction, tvm::runtime::ObjectRef ></a></td><td class="desc"></td></tr>
+<tr id="row_113_76_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::IterVar, tvm::te::IterVarAttr ></a></td><td class="desc"></td></tr>
+<tr id="row_113_77_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::usmp::BufferInfo, tvm::tir::Stmt ></a></td><td class="desc"></td></tr>
+<tr id="row_113_78_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::PrimExpr ></a></td><td class="desc"></td></tr>
+<tr id="row_113_79_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::Range ></a></td><td class="desc"></td></tr>
+<tr id="row_113_80_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::tir::Buffer ></a></td><td class="desc"></td></tr>
+<tr id="row_113_81_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< tvm::tir::Var, tvm::tir::IterVar ></a></td><td class="desc"></td></tr>
+<tr id="row_113_82_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::DiagnosticContext ></a></td><td class="desc"></td></tr>
+<tr id="row_113_83_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::FloatImm ></a></td><td class="desc"></td></tr>
+<tr id="row_113_84_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Integer ></a></td><td class="desc"></td></tr>
+<tr id="row_113_85_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::IRModule ></a></td><td class="desc"></td></tr>
+<tr id="row_113_86_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::CostModel ></a></td><td class="desc"></td></tr>
+<tr id="row_113_87_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::Database ></a></td><td class="desc"></td></tr>
+<tr id="row_113_88_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::SearchStrategy ></a></td><td class="desc"></td></tr>
+<tr id="row_113_89_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::meta_schedule::SpaceGenerator ></a></td><td class="desc"></td></tr>
+<tr id="row_113_90_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::PrimExpr ></a></td><td class="desc"></td></tr>
+<tr id="row_113_91_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::auto_scheduler::MeasureCallback > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_92_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::FloatImm > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_93_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::Integer > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_94_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::ArgInfo > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_95_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::BuilderResult > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_96_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::MeasureCandidate > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_97_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::Postproc > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_98_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::RunnerFuture > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_99_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::meta_schedule::ScheduleRule > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_100_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::tir::BufferRegion > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_101_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Array< tvm::tir::Stmt > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_102_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::meta_schedule::Mutator, tvm::FloatImm > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_103_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::runtime::String, tvm::runtime::NDArray > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_104_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::Map< tvm::runtime::String, tvm::runtime::ObjectRef > ></a></td><td class="desc"></td></tr>
+<tr id="row_113_105_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::NDArray ></a></td><td class="desc"></td></tr>
+<tr id="row_113_106_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::ObjectRef ></a></td><td class="desc"></td></tr>
+<tr id="row_113_107_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::runtime::String ></a></td><td class="desc"></td></tr>
+<tr id="row_113_108_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::script::printer::ExprDoc ></a></td><td class="desc"></td></tr>
+<tr id="row_113_109_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Target ></a></td><td class="desc"></td></tr>
+<tr id="row_113_110_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::tir::IterVar ></a></td><td class="desc"></td></tr>
+<tr id="row_113_111_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::tir::Stmt ></a></td><td class="desc"></td></tr>
+<tr id="row_113_112_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< tvm::Type ></a></td><td class="desc"></td></tr>
+<tr id="row_113_113_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_113_" class="arrow" onclick="toggleFolder('113_113_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1AffineType.html" target="_self">tvm::AffineType</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1AffineTypeNode.html" title="AffineType representation. ">A [...]
+<tr id="row_113_113_0_" 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="classtvm_1_1TensorAffineType.html" target="_self">tvm::TensorAffineType</a></td><td class="desc">Managed reference to AffineTypes </td></tr>
+<tr id="row_113_113_1_" 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="classtvm_1_1TupleAffineType.html" target="_self">tvm::TupleAffineType</a></td><td class="desc">Managed reference to TupleAffineTypes </td></tr>
+<tr id="row_113_114_" 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="classtvm_1_1arith_1_1ConstIntBound.html" target="_self">tvm::arith::ConstIntBound</a></td><td class="desc">Reference class to <a class="el" href="classtvm_1_1arith_1_1ConstIntBoundNode.html" title="Constant integer up and lower bound(inclusive). Useful for value bound analysis. ">Co [...]
+<tr id="row_113_115_" 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="classtvm_1_1arith_1_1IntConstraints.html" target="_self">tvm::arith::IntConstraints</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntConstraintsNode.html" title="Represent integer constrains including (integer) variables, their ranges and the [...]
+<tr id="row_113_116_" 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="classtvm_1_1arith_1_1IntConstraintsTransform.html" target="_self">tvm::arith::IntConstraintsTransform</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntConstraintsTransformNode.html" title="We can have different set of variables to represent t [...]
+<tr id="row_113_117_" 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="classtvm_1_1arith_1_1IntGroupBounds.html" target="_self">tvm::arith::IntGroupBounds</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntGroupBoundsNode.html" title="Represent integer grouped bounds which are classified into lower bounds (inclusi [...]
+<tr id="row_113_118_" 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="classtvm_1_1arith_1_1IntSet.html" target="_self">tvm::arith::IntSet</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IntSetNode.html" title="Base class of all Integer set containers. represent a set of integers in one dimension. ">IntSetNode</a> [...]
+<tr id="row_113_119_" 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="classtvm_1_1arith_1_1IterMapResult.html" target="_self">tvm::arith::IterMapResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterMapResultNode.html" title="Result of DetectIterMap. ">IterMapResultNode</a> </td></tr>
+<tr id="row_113_120_" 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="classtvm_1_1arith_1_1IterMark.html" target="_self">tvm::arith::IterMark</a></td><td class="desc">Managed reference to IterMarkExprNode </td></tr>
+<tr id="row_113_121_" 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="classtvm_1_1arith_1_1ModularSet.html" target="_self">tvm::arith::ModularSet</a></td><td class="desc">Reference of <a class="el" href="classtvm_1_1arith_1_1ModularSetNode.html" title="Range of a linear integer function. Use to do specify the possible index values. ">ModularSetNode</a [...]
+<tr id="row_113_122_" 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="classtvm_1_1AttrFieldInfo.html" target="_self">tvm::AttrFieldInfo</a></td><td class="desc"><a class="el" href="classtvm_1_1AttrFieldInfo.html" title="AttrFieldInfo. ">AttrFieldInfo</a> </td></tr>
+<tr id="row_113_123_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_123_" class="arrow" onclick="toggleFolder('113_123_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Attrs.html" target="_self">tvm::Attrs</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseAttrsNode.html" title="Base class of all attribute class. ">Base [...]
+<tr id="row_113_123_0_" 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="classtvm_1_1DictAttrs.html" target="_self">tvm::DictAttrs</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1DictAttrsNode.html" title="Specialized attribute type that is backed by a map. The DictAttrsNode implements the Attrs behavior...">DictAttrsNode< [...]
+<tr id="row_113_124_" 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="classtvm_1_1auto__scheduler_1_1AccessAnalyzer.html" target="_self">tvm::auto_scheduler::AccessAnalyzer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AccessAnalyzerNode.html" title="Static analyzer for a ComputeDAG. ">AccessAnalyzerN [...]
+<tr id="row_113_125_" 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="classtvm_1_1auto__scheduler_1_1AttachMap.html" target="_self">tvm::auto_scheduler::AttachMap</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AttachMapNode.html" title="stores the compute_at relation between stages This stores a bi-dir [...]
+<tr id="row_113_126_" 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="classtvm_1_1auto__scheduler_1_1BuildResult.html" target="_self">tvm::auto_scheduler::BuildResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1BuildResultNode.html" title="Store the result of a build. ">BuildResultNode</a> </td></tr>
+<tr id="row_113_127_" 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="classtvm_1_1auto__scheduler_1_1ComputeDAG.html" target="_self">tvm::auto_scheduler::ComputeDAG</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeDAGNode.html" title="The auto-scheduler's computational graph and related progra [...]
+<tr id="row_113_128_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_128_" class="arrow" onclick="toggleFolder('113_128_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1CostModel.html" target="_self">tvm::auto_scheduler::CostModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CostMo [...]
+<tr id="row_113_128_0_" 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="classtvm_1_1auto__scheduler_1_1PythonBasedModel.html" target="_self">tvm::auto_scheduler::PythonBasedModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PythonBasedModelNode.html" title="A wrapper for cost model defined by python [...]
+<tr id="row_113_128_1_" 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="classtvm_1_1auto__scheduler_1_1RandomModel.html" target="_self">tvm::auto_scheduler::RandomModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RandomModelNode.html" title="The cost model returning random value for all predictions. [...]
+<tr id="row_113_129_" 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="classtvm_1_1auto__scheduler_1_1HardwareParams.html" target="_self">tvm::auto_scheduler::HardwareParams</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1HardwareParamsNode.html" title="The parameters of target hardware used to guide the [...]
+<tr id="row_113_130_" 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="classtvm_1_1auto__scheduler_1_1Iterator.html" target="_self">tvm::auto_scheduler::Iterator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1IteratorNode.html" title="An iterator of a for-loop Similar to tvm::IterVar in include/tvm/tir/ [...]
+<tr id="row_113_131_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_131_" class="arrow" onclick="toggleFolder('113_131_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureCallback.html" target="_self">tvm::auto_scheduler::MeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__schedul [...]
+<tr id="row_113_131_0_" 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="classtvm_1_1auto__scheduler_1_1PythonBasedMeasureCallback.html" target="_self">tvm::auto_scheduler::PythonBasedMeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PythonBasedMeasureCallbackNode.html" title="A wrapper for [...]
+<tr id="row_113_131_1_" 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="classtvm_1_1auto__scheduler_1_1RecordToFile.html" target="_self">tvm::auto_scheduler::RecordToFile</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RecordToFileNode.html" title="Callback for logging the input and results of measureme [...]
+<tr id="row_113_132_" 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="classtvm_1_1auto__scheduler_1_1MeasureInput.html" target="_self">tvm::auto_scheduler::MeasureInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureInputNode.html" title="Store the input of a measurement. ">MeasureInputNode</a> [...]
+<tr id="row_113_133_" 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="classtvm_1_1auto__scheduler_1_1MeasureResult.html" target="_self">tvm::auto_scheduler::MeasureResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1MeasureResultNode.html" title="Store the results of a measurement. ">MeasureResultNod [...]
+<tr id="row_113_134_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_134_" class="arrow" onclick="toggleFolder('113_134_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramBuilder.html" target="_self">tvm::auto_scheduler::ProgramBuilder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler [...]
+<tr id="row_113_134_0_" 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="classtvm_1_1auto__scheduler_1_1LocalBuilder.html" target="_self">tvm::auto_scheduler::LocalBuilder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1LocalBuilderNode.html" title="LocalBuilder use local CPU cores to build programs in p [...]
+<tr id="row_113_135_" 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="classtvm_1_1auto__scheduler_1_1ProgramMeasurer.html" target="_self">tvm::auto_scheduler::ProgramMeasurer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramMeasurerNode.html" title="Measurer that measures the time costs of tvm pro [...]
+<tr id="row_113_136_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_136_" class="arrow" onclick="toggleFolder('113_136_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1ProgramRunner.html" target="_self">tvm::auto_scheduler::ProgramRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1 [...]
+<tr id="row_113_136_0_" 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="classtvm_1_1auto__scheduler_1_1LocalRunner.html" target="_self">tvm::auto_scheduler::LocalRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1LocalRunnerNode.html" title="LocalRunner that uses local CPU/GPU to measure the time co [...]
+<tr id="row_113_136_1_" 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="classtvm_1_1auto__scheduler_1_1RPCRunner.html" target="_self">tvm::auto_scheduler::RPCRunner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RPCRunnerNode.html" title="RPCRunner that uses RPC call to measures the time cost of progra [...]
+<tr id="row_113_137_" 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="classtvm_1_1auto__scheduler_1_1RecordReader.html" target="_self">tvm::auto_scheduler::RecordReader</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RecordReaderNode.html" title="Log reader to load step logs from a file. ">RecordReaderN [...]
+<tr id="row_113_138_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_138_" class="arrow" onclick="toggleFolder('113_138_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1SearchCallback.html" target="_self">tvm::auto_scheduler::SearchCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler [...]
+<tr id="row_113_138_0_" 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="classtvm_1_1auto__scheduler_1_1PreloadMeasuredStates.html" target="_self">tvm::auto_scheduler::PreloadMeasuredStates</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PreloadMeasuredStatesNode.html" title="Preload measured states from [...]
+<tr id="row_113_139_" 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="classtvm_1_1auto__scheduler_1_1SearchPolicy.html" target="_self">tvm::auto_scheduler::SearchPolicy</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchPolicyNode.html" title="The base class of search policies. ">SearchPolicyNode</a> [...]
+<tr id="row_113_140_" 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="classtvm_1_1auto__scheduler_1_1SearchTask.html" target="_self">tvm::auto_scheduler::SearchTask</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SearchTaskNode.html" title="The computation information and hardware parameters for a speci [...]
+<tr id="row_113_141_" 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="classtvm_1_1auto__scheduler_1_1Stage.html" target="_self">tvm::auto_scheduler::Stage</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StageNode.html" title="A op stage in the compute declaration. Similar to te::Stage in include/tvm/te/ [...]
+<tr id="row_113_142_" 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="classtvm_1_1auto__scheduler_1_1State.html" target="_self">tvm::auto_scheduler::State</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StateNode.html" title="A state in the search process. It consists of the current loop structure and a [...]
+<tr id="row_113_143_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_143_" class="arrow" onclick="toggleFolder('113_143_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1auto__scheduler_1_1Step.html" target="_self">tvm::auto_scheduler::Step</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StepNode.html" t [...]
+<tr id="row_113_143_0_" 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="classtvm_1_1auto__scheduler_1_1AnnotationStep.html" target="_self">tvm::auto_scheduler::AnnotationStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1AnnotationStepNode.html" title="Annotation step that corresponds to vectorize, pa [...]
+<tr id="row_113_143_1_" 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="classtvm_1_1auto__scheduler_1_1CacheReadStep.html" target="_self">tvm::auto_scheduler::CacheReadStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CacheReadStepNode.html" title="Cache read step that corresponds to te::Schedule::ca [...]
+<tr id="row_113_143_2_" 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="classtvm_1_1auto__scheduler_1_1CacheWriteStep.html" target="_self">tvm::auto_scheduler::CacheWriteStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1CacheWriteStepNode.html" title="Cache write step that corresponds to te::Schedule [...]
+<tr id="row_113_143_3_" 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="classtvm_1_1auto__scheduler_1_1ComputeAtStep.html" target="_self">tvm::auto_scheduler::ComputeAtStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeAtStepNode.html" title="Compute at step that corresponds to te::Stage::compu [...]
+<tr id="row_113_143_4_" 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="classtvm_1_1auto__scheduler_1_1ComputeInlineStep.html" target="_self">tvm::auto_scheduler::ComputeInlineStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeInlineStepNode.html" title="Compute inline step that corresponds to [...]
+<tr id="row_113_143_5_" 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="classtvm_1_1auto__scheduler_1_1ComputeRootStep.html" target="_self">tvm::auto_scheduler::ComputeRootStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ComputeRootStepNode.html" title="Compute root step that corresponds to te::Stag [...]
+<tr id="row_113_143_6_" 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="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStep.html" target="_self">tvm::auto_scheduler::FollowFusedSplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FollowFusedSplitStepNode.html" title="Similar to FollowSplitStep, but [...]
+<tr id="row_113_143_7_" 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="classtvm_1_1auto__scheduler_1_1FollowSplitStep.html" target="_self">tvm::auto_scheduler::FollowSplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FollowSplitStepNode.html" title="Similar to SplitStepNode, but uses split facto [...]
+<tr id="row_113_143_8_" 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="classtvm_1_1auto__scheduler_1_1FuseStep.html" target="_self">tvm::auto_scheduler::FuseStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1FuseStepNode.html" title="Fuse step that corresponds to te::Stage::fuse. ">FuseStepNode</a> < [...]
+<tr id="row_113_143_9_" 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="classtvm_1_1auto__scheduler_1_1PragmaStep.html" target="_self">tvm::auto_scheduler::PragmaStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1PragmaStepNode.html" title="Pragma step that corresponds to te::Stage::pragma. ">PragmaSt [...]
+<tr id="row_113_143_10_" 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="classtvm_1_1auto__scheduler_1_1ReorderStep.html" target="_self">tvm::auto_scheduler::ReorderStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1ReorderStepNode.html" title="Reorder step that corresponds to te::Stage::reorder. ">Re [...]
+<tr id="row_113_143_11_" 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="classtvm_1_1auto__scheduler_1_1RfactorStep.html" target="_self">tvm::auto_scheduler::RfactorStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1RfactorStepNode.html" title="Reduction factor step that corresponds to te::Schedule::r [...]
+<tr id="row_113_143_12_" 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="classtvm_1_1auto__scheduler_1_1SplitStep.html" target="_self">tvm::auto_scheduler::SplitStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1SplitStepNode.html" title="Split step that corresponds to te::Stage::split with additional [...]
+<tr id="row_113_143_13_" 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="classtvm_1_1auto__scheduler_1_1StorageAlignStep.html" target="_self">tvm::auto_scheduler::StorageAlignStep</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1StorageAlignStepNode.html" title="Storage align step that corresponds to te: [...]
+<tr id="row_113_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="classtvm_1_1auto__scheduler_1_1TuningOptions.html" target="_self">tvm::auto_scheduler::TuningOptions</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1auto__scheduler_1_1TuningOptionsNode.html" title="Tuning and measurement options. ">TuningOptionsNode</a [...]
+<tr id="row_113_145_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_145_" class="arrow" onclick="toggleFolder('113_145_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1BaseExpr.html" target="_self">tvm::BaseExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseExprNode.html" title="Base type of all the expressions. "> [...]
+<tr id="row_113_145_0_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_145_0_" class="arrow" onclick="toggleFolder('113_145_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1PrimExpr.html" target="_self">tvm::PrimExpr</a></td><td class="desc">Reference to <a class="el" href="classtvm_1_1PrimExprNode.html" title="Base node of all primitive expressions [...]
+<tr id="row_113_145_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_145_0_0_" class="arrow" onclick="toggleFolder('113_145_0_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterMapExpr.html" target="_self">tvm::arith::IterMapExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterMapExprNode.htm [...]
+<tr id="row_113_145_0_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterSplitExpr.html" target="_self">tvm::arith::IterSplitExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterSplitExprNode.html" title="Split of an iterator. ">IterSplitExprNode</a> </td></tr>
+<tr id="row_113_145_0_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1arith_1_1IterSumExpr.html" target="_self">tvm::arith::IterSumExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1arith_1_1IterSumExprNode.html" title="Fuse multiple iterators by summing them with scaling. ">IterSumExprNode</a> </td></tr>
+<tr id="row_113_145_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_1FloatImm.html" target="_self">tvm::FloatImm</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1FloatImmNode.html" title="Constant floating point literals in the program. ">FloatImmNode</a> </td></tr>
+<tr id="row_113_145_0_2_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_145_0_2_" class="arrow" onclick="toggleFolder('113_145_0_2_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1IntImm.html" target="_self">tvm::IntImm</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1IntImmNode.html" title="Constant integer literal [...]
+<tr id="row_113_145_0_2_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Bool.html" target="_self">tvm::Bool</a></td><td class="desc">Boolean constant </td></tr>
+<tr id="row_113_145_0_2_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1Integer.html" target="_self">tvm::Integer</a></td><td class="desc">Container of constant int that adds more constructors </td></tr>
+<tr id="row_113_145_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_1tir_1_1Add.html" target="_self">tvm::tir::Add</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AddNode.html" title="a + b ">AddNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1And.html" target="_self">tvm::tir::And</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AndNode.html" title="a && b ">AndNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1Any.html" target="_self">tvm::tir::Any</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1AnyNode.html" title="Any shape. ">AnyNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1Broadcast.html" target="_self">tvm::tir::Broadcast</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BroadcastNode.html" title="Create a vector where all the elements are value. ">BroadcastNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1BufferLoad.html" target="_self">tvm::tir::BufferLoad</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferLoadNode.html" title="Load value from the high dimension buffer. ">BufferLoadNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1Call.html" target="_self">tvm::tir::Call</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CallNode.html" title="Call node. ">CallNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1Cast.html" target="_self">tvm::tir::Cast</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CastNode.html" title="Cast value from one data type to another. ">CastNode</a> </td></tr>
+<tr id="row_113_145_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_1tir_1_1Div.html" target="_self">tvm::tir::Div</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DivNode.html" title="a / b in the C semnatics. ">DivNode</a> </td></tr>
+<tr id="row_113_145_0_11_" 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_1tir_1_1EQ.html" target="_self">tvm::tir::EQ</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1EQNode.html" title="a == b ">EQNode</a> </td></tr>
+<tr id="row_113_145_0_12_" 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_1tir_1_1FloorDiv.html" target="_self">tvm::tir::FloorDiv</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1FloorDivNode.html" title="Floor division, floor(a/b) ">FloorDivNode</a> </td></tr>
+<tr id="row_113_145_0_13_" 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_1tir_1_1FloorMod.html" target="_self">tvm::tir::FloorMod</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1FloorModNode.html" title="The remainder of the floordiv. ">FloorModNode</a> </td></tr>
+<tr id="row_113_145_0_14_" 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_1tir_1_1GE.html" target="_self">tvm::tir::GE</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1GENode.html" title="a >= b ">GENode</a> </td></tr>
+<tr id="row_113_145_0_15_" 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_1tir_1_1GT.html" target="_self">tvm::tir::GT</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1GTNode.html" title="a > b ">GTNode</a> </td></tr>
+<tr id="row_113_145_0_16_" 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_1tir_1_1LE.html" target="_self">tvm::tir::LE</a></td><td class="desc">Managed reference to <a class="el" href="structtvm_1_1tir_1_1LENode.html" title="a <= b ">LENode</a> </td></tr>
+<tr id="row_113_145_0_17_" 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_1tir_1_1Let.html" target="_self">tvm::tir::Let</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LetNode.html" title="Let binding. Bind var to value then evaluate body. ">LetNode</a> </td></tr>
+<tr id="row_113_145_0_18_" 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_1tir_1_1Load.html" target="_self">tvm::tir::Load</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LoadNode.html" title="Load the value from buffer_var. ">LoadNode</a> </td></tr>
+<tr id="row_113_145_0_19_" 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_1tir_1_1LT.html" target="_self">tvm::tir::LT</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LTNode.html" title="a < b ">LTNode</a> </td></tr>
+<tr id="row_113_145_0_20_" 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_1tir_1_1Max.html" target="_self">tvm::tir::Max</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MaxNode.html" title="max(a, b) ">MaxNode</a> </td></tr>
+<tr id="row_113_145_0_21_" 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_1tir_1_1Min.html" target="_self">tvm::tir::Min</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MinNode.html" title="min(a, b) ">MinNode</a> </td></tr>
+<tr id="row_113_145_0_22_" 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_1tir_1_1Mod.html" target="_self">tvm::tir::Mod</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ModNode.html" title="a % b in the C semnatics. ">ModNode</a> </td></tr>
+<tr id="row_113_145_0_23_" 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_1tir_1_1Mul.html" target="_self">tvm::tir::Mul</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MulNode.html" title="a * b ">MulNode</a> </td></tr>
+<tr id="row_113_145_0_24_" 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_1tir_1_1NE.html" target="_self">tvm::tir::NE</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1NENode.html" title="a != b ">NENode</a> </td></tr>
+<tr id="row_113_145_0_25_" 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_1tir_1_1Not.html" target="_self">tvm::tir::Not</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1NotNode.html" title="!a ">NotNode</a> </td></tr>
+<tr id="row_113_145_0_26_" 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_1tir_1_1Or.html" target="_self">tvm::tir::Or</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1OrNode.html" title="a || b ">OrNode</a> </td></tr>
+<tr id="row_113_145_0_27_" 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_1tir_1_1ProducerLoad.html" target="_self">tvm::tir::ProducerLoad</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ProducerLoadNode.html" title="Load value from the result produced by the producer. ">ProducerLoadNode</a> </td></tr>
+<tr id="row_113_145_0_28_" 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_1tir_1_1Ramp.html" target="_self">tvm::tir::Ramp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1RampNode.html" title="Construct a vector with lanes elements where its i-th element equals base + i * stride. This is useful to const [...]
+<tr id="row_113_145_0_29_" 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_1tir_1_1Reduce.html" target="_self">tvm::tir::Reduce</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ReduceNode.html" title="Reduction operator operator. ">ReduceNode</a> </td></tr>
+<tr id="row_113_145_0_30_" 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_1tir_1_1Select.html" target="_self">tvm::tir::Select</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1SelectNode.html" title="return true_value if condition is true, otherwise return false_value. ">SelectNode</a> </td></tr>
+<tr id="row_113_145_0_31_" 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_1tir_1_1Shuffle.html" target="_self">tvm::tir::Shuffle</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1ShuffleNode.html" title="Shuffle instruction. vec = concat(vectors) result = (vec[indices[0]], vec[indices[1]] ...">ShuffleNode [...]
+<tr id="row_113_145_0_32_" 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_1tir_1_1StringImm.html" target="_self">tvm::tir::StringImm</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1StringImmNode.html" title="String constants, only used in asserts. ">StringImmNode</a> </td></tr>
+<tr id="row_113_145_0_33_" 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_1tir_1_1Sub.html" target="_self">tvm::tir::Sub</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1SubNode.html" title="a - b ">SubNode</a> </td></tr>
+<tr id="row_113_145_0_34_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_145_0_34_" class="arrow" onclick="toggleFolder('113_145_0_34_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1Var.html" target="_self">tvm::tir::Var</a></td><td class="desc">Named variable in TIR </td></tr>
+<tr id="row_113_145_0_34_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1SizeVar.html" target="_self">tvm::tir::SizeVar</a></td><td class="desc">Named variable represents a tensor index size </td></tr>
+<tr id="row_113_145_1_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_145_1_" class="arrow" onclick="toggleFolder('113_145_1_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1RelayExpr.html" target="_self">tvm::RelayExpr</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1RelayExprNode.html" title="Base node of all non-primit [...]
+<tr id="row_113_145_1_0_" class="even" style="display:none;"><td class="entry"><span style="width:48px;display:inline-block;"> </span><span id="arr_113_145_1_0_" class="arrow" onclick="toggleFolder('113_145_1_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1BaseFunc.html" target="_self">tvm::BaseFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1BaseFuncNode.html" title="Base node of all functio [...]
+<tr id="row_113_145_1_0_0_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1Function.html" target="_self">tvm::relay::Function</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1relay_1_1FunctionNode.html" title="Relay Function container. ">FunctionNode</a> </td></tr>
+<tr id="row_113_145_1_0_1_" class="even" style="display:none;"><td class="entry"><span style="width:80px;display:inline-block;"> </span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1PrimFunc.html" target="_self">tvm::tir::PrimFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1PrimFuncNode.html" title="Primitive functions that contains TIR statements. ">PrimFuncNode</a> </td></tr>
+<tr id="row_113_145_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_1Constructor.html" target="_self">tvm::Constructor</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1ConstructorNode.html" title="ADT constructor. Constructors compare by pointer equality. ">ConstructorNode</a> </td></tr>
+<tr id="row_113_145_1_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_1GlobalVar.html" target="_self">tvm::GlobalVar</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1GlobalVarNode.html" title="Global variable that lives in the top-level module. ">GlobalVarNode</a> </td></tr>
+<tr id="row_113_145_1_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_1Op.html" target="_self">tvm::Op</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1OpNode.html" title="Primitive Op(builtin intrinsics) ">OpNode</a> </td></tr>
+<tr id="row_113_145_1_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_1relay_1_1Call.html" target="_self">tvm::relay::Call</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1Constant.html" target="_self">tvm::relay::Constant</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1If.html" target="_self">tvm::relay::If</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1Let.html" target="_self">tvm::relay::Let</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1Match.html" target="_self">tvm::relay::Match</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1RefCreate.html" target="_self">tvm::relay::RefCreate</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_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_1relay_1_1RefRead.html" target="_self">tvm::relay::RefRead</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_11_" 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_1relay_1_1RefWrite.html" target="_self">tvm::relay::RefWrite</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_12_" 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_1relay_1_1TempExpr.html" target="_self">tvm::relay::TempExpr</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_13_" 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_1relay_1_1Tuple.html" target="_self">tvm::relay::Tuple</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_14_" 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_1relay_1_1TupleGetItem.html" target="_self">tvm::relay::TupleGetItem</a></td><td class="desc"></td></tr>
+<tr id="row_113_145_1_15_" 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_1relay_1_1Var.html" target="_self">tvm::relay::Var</a></td><td class="desc"></td></tr>
+<tr id="row_113_146_" 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="classtvm_1_1CompilationConfig.html" target="_self">tvm::CompilationConfig</a></td><td class="desc">Managed reference class to <code><a class="el" href="classtvm_1_1CompilationConfig.html" title="Managed reference class to CompilationConfig. ">CompilationConfig</a></code> </td></tr>
+<tr id="row_113_147_" 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="classtvm_1_1ConstantInfo.html" target="_self">tvm::ConstantInfo</a></td><td class="desc"></td></tr>
+<tr id="row_113_148_" 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="classtvm_1_1ConstantMemoryPools.html" target="_self">tvm::ConstantMemoryPools</a></td><td class="desc"></td></tr>
+<tr id="row_113_149_" 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="classtvm_1_1Diagnostic.html" target="_self">tvm::Diagnostic</a></td><td class="desc"></td></tr>
+<tr id="row_113_150_" 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="classtvm_1_1DiagnosticContext.html" target="_self">tvm::DiagnosticContext</a></td><td class="desc"></td></tr>
+<tr id="row_113_151_" 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="classtvm_1_1DiagnosticRenderer.html" target="_self">tvm::DiagnosticRenderer</a></td><td class="desc"></td></tr>
+<tr id="row_113_152_" 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="classtvm_1_1EnvFunc.html" target="_self">tvm::EnvFunc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1EnvFuncNode.html" title="A serializable function backed by TVM's global environment. ">EnvFuncNode</a> </td></tr>
+<tr id="row_113_153_" 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="classtvm_1_1GenericFunc.html" target="_self">tvm::GenericFunc</a></td><td class="desc">Generic function that can be specialized on a per-target basis </td></tr>
+<tr id="row_113_154_" 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="classtvm_1_1GlobalVarSupply.html" target="_self">tvm::GlobalVarSupply</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1GlobalVarSupplyNode.html" title="GlobalVarSupply can be used to generate unique GlobalVars. ">GlobalVarSupplyNode</a> </td></tr>
+<tr id="row_113_155_" 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="classtvm_1_1instrument_1_1PassInstrument.html" target="_self">tvm::instrument::PassInstrument</a></td><td class="desc">Managed reference class for <a class="el" href="classtvm_1_1instrument_1_1PassInstrumentNode.html" title="PassInstrumentNode forms an instrument implementation. It [...]
+<tr id="row_113_156_" 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="classtvm_1_1IRModule.html" target="_self">tvm::IRModule</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1IRModuleNode.html" title="IRModule that holds functions and type definitions. ">IRModuleNode</a> </td></tr>
+<tr id="row_113_157_" 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="classtvm_1_1MemoryInfo.html" target="_self">tvm::MemoryInfo</a></td><td class="desc">Defines memory info </td></tr>
+<tr id="row_113_158_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_158_" class="arrow" onclick="toggleFolder('113_158_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfo.html" target="_self">tvm::meta_schedule::ArgInfo</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ArgInfoNode.h [...]
+<tr id="row_113_158_0_" 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="classtvm_1_1meta__schedule_1_1TensorInfo.html" target="_self">tvm::meta_schedule::TensorInfo</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TensorInfoNode.html" title="The tensor argument information. ">TensorInfoNode</a> </td></tr>
+<tr id="row_113_159_" 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="classtvm_1_1meta__schedule_1_1Builder.html" target="_self">tvm::meta_schedule::Builder</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderNode.html" title="The abstract builder interface. ">BuilderNode</a> </td></tr>
+<tr id="row_113_160_" 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="classtvm_1_1meta__schedule_1_1BuilderInput.html" target="_self">tvm::meta_schedule::BuilderInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderInputNode.html" title="The builder's input, containing an IRModule and the targ [...]
+<tr id="row_113_161_" 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="classtvm_1_1meta__schedule_1_1BuilderResult.html" target="_self">tvm::meta_schedule::BuilderResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1BuilderResultNode.html" title="The builder's output, containing the artifact path or [...]
+<tr id="row_113_162_" 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="classtvm_1_1meta__schedule_1_1CostModel.html" target="_self">tvm::meta_schedule::CostModel</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1CostModelNode.html" title="Cost model. ">CostModelNode</a> </td></tr>
+<tr id="row_113_163_" 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="classtvm_1_1meta__schedule_1_1Database.html" target="_self">tvm::meta_schedule::Database</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1DatabaseNode.html">DatabaseNode</a> </td></tr>
+<tr id="row_113_164_" 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="classtvm_1_1meta__schedule_1_1ExtractedTask.html" target="_self">tvm::meta_schedule::ExtractedTask</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ExtractedTaskNode.html" title="A tuning task extracted from the high-level IR. ">Extract [...]
+<tr id="row_113_165_" 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="classtvm_1_1meta__schedule_1_1FeatureExtractor.html" target="_self">tvm::meta_schedule::FeatureExtractor</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1FeatureExtractorNode.html" title="Extractor for features from measure candidates f [...]
+<tr id="row_113_166_" 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="classtvm_1_1meta__schedule_1_1MeasureCallback.html" target="_self">tvm::meta_schedule::MeasureCallback</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCallbackNode.html" title="Rules to apply after measure results is available. [...]
+<tr id="row_113_167_" 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="classtvm_1_1meta__schedule_1_1MeasureCandidate.html" target="_self">tvm::meta_schedule::MeasureCandidate</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MeasureCandidateNode.html" title="The schedule (with input shapes) to be measured. [...]
+<tr id="row_113_168_" 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="classtvm_1_1meta__schedule_1_1Mutator.html" target="_self">tvm::meta_schedule::Mutator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1MutatorNode.html" title="Mutator is designed to mutate the trace to explore the design space. ">Muta [...]
+<tr id="row_113_169_" 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="classtvm_1_1meta__schedule_1_1Postproc.html" target="_self">tvm::meta_schedule::Postproc</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1PostprocNode.html" title="Rules to apply a postprocessor to a schedule. ">PostprocNode</a> </td></tr>
+<tr id="row_113_170_" 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="classtvm_1_1meta__schedule_1_1Profiler.html" target="_self">tvm::meta_schedule::Profiler</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ProfilerNode.html" title="A generic profiler. ">ProfilerNode</a> </td></tr>
+<tr id="row_113_171_" 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="classtvm_1_1meta__schedule_1_1Runner.html" target="_self">tvm::meta_schedule::Runner</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerNode.html" title="The abstract runner interface. ">RunnerNode</a> </td></tr>
+<tr id="row_113_172_" 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="classtvm_1_1meta__schedule_1_1RunnerFuture.html" target="_self">tvm::meta_schedule::RunnerFuture</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerFutureNode.html" title="A class to asynchronously fetch runner's output. ">Runne [...]
+<tr id="row_113_173_" 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="classtvm_1_1meta__schedule_1_1RunnerInput.html" target="_self">tvm::meta_schedule::RunnerInput</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerInputNode.html" title="Runner's input containing path of artifact, type of device [...]
+<tr id="row_113_174_" 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="classtvm_1_1meta__schedule_1_1RunnerResult.html" target="_self">tvm::meta_schedule::RunnerResult</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1RunnerResultNode.html" title="Runner's output containing measurement result of Measure [...]
+<tr id="row_113_175_" 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="classtvm_1_1meta__schedule_1_1ScheduleRule.html" target="_self">tvm::meta_schedule::ScheduleRule</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1ScheduleRuleNode.html" title="Rules to modify a block in a schedule. ">ScheduleRuleNode</a [...]
+<tr id="row_113_176_" 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="classtvm_1_1meta__schedule_1_1SearchStrategy.html" target="_self">tvm::meta_schedule::SearchStrategy</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1SearchStrategyNode.html" title="The search strategy for measure candidates generation. [...]
+<tr id="row_113_177_" 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="classtvm_1_1meta__schedule_1_1SpaceGenerator.html" target="_self">tvm::meta_schedule::SpaceGenerator</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1SpaceGeneratorNode.html" title="The abstract class for design space generation. ">Spac [...]
+<tr id="row_113_178_" 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="classtvm_1_1meta__schedule_1_1TaskRecord.html" target="_self">tvm::meta_schedule::TaskRecord</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TaskRecordNode.html">TaskRecordNode</a> </td></tr>
+<tr id="row_113_179_" 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="classtvm_1_1meta__schedule_1_1TaskScheduler.html" target="_self">tvm::meta_schedule::TaskScheduler</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TaskSchedulerNode.html" title="The abstract interface of task schedulers. ">TaskSchedule [...]
+<tr id="row_113_180_" 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="classtvm_1_1meta__schedule_1_1TuneContext.html" target="_self">tvm::meta_schedule::TuneContext</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1TuneContextNode.html" title="The auto tuning context. ">TuneContextNode</a> </td></tr>
+<tr id="row_113_181_" 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="classtvm_1_1meta__schedule_1_1TuningRecord.html" target="_self">tvm::meta_schedule::TuningRecord</a></td><td class="desc">The managed reference of <a class="el" href="classtvm_1_1meta__schedule_1_1TuningRecordNode.html" title="The class of tuning records. ">TuningRecordNode</a> </td></tr>
+<tr id="row_113_182_" 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="classtvm_1_1meta__schedule_1_1Workload.html" target="_self">tvm::meta_schedule::Workload</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1meta__schedule_1_1WorkloadNode.html" title="A workload, i.e. an IRModule and its structural hash. ">WorkloadNode</a> [...]
+<tr id="row_113_183_" 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="classtvm_1_1NameSupply.html" target="_self">tvm::NameSupply</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1NameSupplyNode.html" title="NameSupply can be used to generate unique names. ">NameSupplyNode</a> </td></tr>
+<tr id="row_113_184_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_184_" class="arrow" onclick="toggleFolder('113_184_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1ObjectPath.html" target="_self">tvm::ObjectPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_0_" 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="classtvm_1_1ArrayIndexPath.html" target="_self">tvm::ArrayIndexPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_1_" 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="classtvm_1_1AttributeAccessPath.html" target="_self">tvm::AttributeAccessPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_2_" 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="classtvm_1_1MapValuePath.html" target="_self">tvm::MapValuePath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_3_" 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="classtvm_1_1MissingArrayElementPath.html" target="_self">tvm::MissingArrayElementPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_4_" 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="classtvm_1_1MissingMapEntryPath.html" target="_self">tvm::MissingMapEntryPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_5_" 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="classtvm_1_1RootPath.html" target="_self">tvm::RootPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_184_6_" 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="classtvm_1_1UnknownAttributeAccessPath.html" target="_self">tvm::UnknownAttributeAccessPath</a></td><td class="desc"></td></tr>
+<tr id="row_113_185_" 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="classtvm_1_1ObjectPathPair.html" target="_self">tvm::ObjectPathPair</a></td><td class="desc"></td></tr>
+<tr id="row_113_186_" 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="classtvm_1_1parser_1_1Source.html" target="_self">tvm::parser::Source</a></td><td class="desc"></td></tr>
+<tr id="row_113_187_" 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="classtvm_1_1parser_1_1SourceMap.html" target="_self">tvm::parser::SourceMap</a></td><td class="desc"></td></tr>
+<tr id="row_113_188_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_188_" class="arrow" onclick="toggleFolder('113_188_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1PoolInfo.html" target="_self">tvm::PoolInfo</a></td><td class="desc">Base class for <a class="el" href="classtvm_1_1WorkspacePoolInfo.html">WorkspacePoolInfo</a> and <a class="el" href [...]
+<tr id="row_113_188_0_" 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="classtvm_1_1ConstantPoolInfo.html" target="_self">tvm::ConstantPoolInfo</a></td><td class="desc"></td></tr>
+<tr id="row_113_188_1_" 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="classtvm_1_1WorkspacePoolInfo.html" target="_self">tvm::WorkspacePoolInfo</a></td><td class="desc"></td></tr>
+<tr id="row_113_189_" 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="classtvm_1_1PoolInfoProperties.html" target="_self">tvm::PoolInfoProperties</a></td><td class="desc"></td></tr>
+<tr id="row_113_190_" 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="classtvm_1_1Range.html" target="_self">tvm::Range</a></td><td class="desc"><a class="el" href="classtvm_1_1Range.html" title="Range constainer. ">Range</a> constainer </td></tr>
+<tr id="row_113_191_" 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="classtvm_1_1relay_1_1Clause.html" target="_self">tvm::relay::Clause</a></td><td class="desc"></td></tr>
+<tr id="row_113_192_" 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="classtvm_1_1relay_1_1ConstructorValue.html" target="_self">tvm::relay::ConstructorValue</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_193_" class="arrow" onclick="toggleFolder('113_193_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1DFPattern.html" target="_self">tvm::relay::DFPattern</a></td><td class="desc">Managed reference to dataflow patterns </td></tr>
+<tr id="row_113_193_0_" 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="classtvm_1_1relay_1_1AltPattern.html" target="_self">tvm::relay::AltPattern</a></td><td class="desc">A pattern which matches either of two patterns </td></tr>
+<tr id="row_113_193_1_" 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="classtvm_1_1relay_1_1AttrPattern.html" target="_self">tvm::relay::AttrPattern</a></td><td class="desc">A pattern which matches attributes in another pattern </td></tr>
+<tr id="row_113_193_2_" 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="classtvm_1_1relay_1_1CallPattern.html" target="_self">tvm::relay::CallPattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_3_" 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="classtvm_1_1relay_1_1ConstantPattern.html" target="_self">tvm::relay::ConstantPattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_4_" 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="classtvm_1_1relay_1_1DataTypePattern.html" target="_self">tvm::relay::DataTypePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
+<tr id="row_113_193_5_" 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="classtvm_1_1relay_1_1DominatorPattern.html" target="_self">tvm::relay::DominatorPattern</a></td><td class="desc">A pattern which matches a variable length dominator path </td></tr>
+<tr id="row_113_193_6_" 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="classtvm_1_1relay_1_1ExprPattern.html" target="_self">tvm::relay::ExprPattern</a></td><td class="desc">A pattern which matches a literal expression </td></tr>
+<tr id="row_113_193_7_" 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="classtvm_1_1relay_1_1FunctionPattern.html" target="_self">tvm::relay::FunctionPattern</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1relay_1_1FunctionNode.html" title="Relay Function container. ">FunctionNode</a> </td></tr>
+<tr id="row_113_193_8_" 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="classtvm_1_1relay_1_1IfPattern.html" target="_self">tvm::relay::IfPattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_9_" 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="classtvm_1_1relay_1_1LetPattern.html" target="_self">tvm::relay::LetPattern</a></td><td class="desc"><a class="el" href="classtvm_1_1relay_1_1Let.html">Let</a> binding that binds a local var </td></tr>
+<tr id="row_113_193_10_" 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="classtvm_1_1relay_1_1ShapePattern.html" target="_self">tvm::relay::ShapePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
+<tr id="row_113_193_11_" 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="classtvm_1_1relay_1_1TupleGetItemPattern.html" target="_self">tvm::relay::TupleGetItemPattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_12_" 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="classtvm_1_1relay_1_1TuplePattern.html" target="_self">tvm::relay::TuplePattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_13_" 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="classtvm_1_1relay_1_1TypePattern.html" target="_self">tvm::relay::TypePattern</a></td><td class="desc">A pattern which matches a type in another pattern </td></tr>
+<tr id="row_113_193_14_" 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="classtvm_1_1relay_1_1VarPattern.html" target="_self">tvm::relay::VarPattern</a></td><td class="desc"></td></tr>
+<tr id="row_113_193_15_" 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="classtvm_1_1relay_1_1WildcardPattern.html" target="_self">tvm::relay::WildcardPattern</a></td><td class="desc">A pattern which matches anything </td></tr>
+<tr id="row_113_194_" 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="classtvm_1_1relay_1_1DFPatternCallback.html" target="_self">tvm::relay::DFPatternCallback</a></td><td class="desc">Managed reference to dataflow pattern callbacks </td></tr>
+<tr id="row_113_195_" 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="classtvm_1_1relay_1_1Executor.html" target="_self">tvm::relay::Executor</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1relay_1_1ExecutorNode.html" title="Executor information. ">ExecutorNode</a> </td></tr>
+<tr id="row_113_196_" 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="classtvm_1_1relay_1_1Id.html" target="_self">tvm::relay::Id</a></td><td class="desc"></td></tr>
+<tr id="row_113_197_" 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="classtvm_1_1relay_1_1OpImplementation.html" target="_self">tvm::relay::OpImplementation</a></td><td class="desc">Operator implementation class </td></tr>
+<tr id="row_113_198_" 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="classtvm_1_1relay_1_1OpSpecialization.html" target="_self">tvm::relay::OpSpecialization</a></td><td class="desc">Operator specialization class </td></tr>
+<tr id="row_113_199_" 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="classtvm_1_1relay_1_1OpStrategy.html" target="_self">tvm::relay::OpStrategy</a></td><td class="desc">Operator strategy class </td></tr>
+<tr id="row_113_200_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_200_" class="arrow" onclick="toggleFolder('113_200_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1relay_1_1Pattern.html" target="_self">tvm::relay::Pattern</a></td><td class="desc"><a class="el" href="classtvm_1_1relay_1_1Pattern.html" title="Pattern is the base type for an ADT mat [...]
+<tr id="row_113_200_0_" 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="classtvm_1_1relay_1_1PatternConstructor.html" target="_self">tvm::relay::PatternConstructor</a></td><td class="desc"></td></tr>
+<tr id="row_113_200_1_" 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="classtvm_1_1relay_1_1PatternTuple.html" target="_self">tvm::relay::PatternTuple</a></td><td class="desc"></td></tr>
+<tr id="row_113_200_2_" 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="classtvm_1_1relay_1_1PatternVar.html" target="_self">tvm::relay::PatternVar</a></td><td class="desc"></td></tr>
+<tr id="row_113_200_3_" 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="classtvm_1_1relay_1_1PatternWildcard.html" target="_self">tvm::relay::PatternWildcard</a></td><td class="desc"></td></tr>
+<tr id="row_113_201_" 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="classtvm_1_1relay_1_1RecClosure.html" target="_self">tvm::relay::RecClosure</a></td><td class="desc"></td></tr>
+<tr id="row_113_202_" 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="classtvm_1_1relay_1_1RefValue.html" target="_self">tvm::relay::RefValue</a></td><td class="desc"></td></tr>
+<tr id="row_113_203_" 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="classtvm_1_1relay_1_1Runtime.html" target="_self">tvm::relay::Runtime</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1relay_1_1RuntimeNode.html" title="Runtime information. ">RuntimeNode</a> </td></tr>
+<tr id="row_113_204_" 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="classtvm_1_1runtime_1_1ADT.html" target="_self">tvm::runtime::ADT</a></td><td class="desc">Reference to algebraic data type objects </td></tr>
+<tr id="row_113_205_" 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="classtvm_1_1runtime_1_1Array.html" target="_self">tvm::runtime::Array< T, typename ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Array.html" title="Array, container representing a contiguous sequence of ObjectRefs. ">Array</a>, container representing a [...]
+<tr id="row_113_206_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_206_" class="arrow" onclick="toggleFolder('113_206_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1Closure.html" target="_self">tvm::runtime::Closure</a></td><td class="desc">Reference to closure </td></tr>
+<tr id="row_113_206_0_" 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="classtvm_1_1relay_1_1InterpreterClosure.html" target="_self">tvm::relay::InterpreterClosure</a></td><td class="desc"></td></tr>
+<tr id="row_113_206_1_" 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="classtvm_1_1runtime_1_1vm_1_1VMClosure.html" target="_self">tvm::runtime::vm::VMClosure</a></td><td class="desc">Reference to closure </td></tr>
+<tr id="row_113_207_" 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="classtvm_1_1runtime_1_1Map.html" target="_self">tvm::runtime::Map< K, V, typename, typename ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Map.html" title="Map container of NodeRef->NodeRef in DSL graph. Map implements copy on write semantics, which m [...]
+<tr id="row_113_208_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_208_" class="arrow" onclick="toggleFolder('113_208_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1runtime_1_1metadata_1_1MetadataBase.html" target="_self">tvm::runtime::metadata::MetadataBase</a></td><td class="desc">Reference class for the common <a class="el" href="classtvm_1_1ru [...]
+<tr id="row_113_208_0_" 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="classtvm_1_1runtime_1_1metadata_1_1ConstantInfoMetadata.html" target="_self">tvm::runtime::metadata::ConstantInfoMetadata</a></td><td class="desc"></td></tr>
+<tr id="row_113_208_1_" 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="classtvm_1_1runtime_1_1metadata_1_1Metadata.html" target="_self">tvm::runtime::metadata::Metadata</a></td><td class="desc"></td></tr>
+<tr id="row_113_208_2_" 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="classtvm_1_1runtime_1_1metadata_1_1MetadataArray.html" target="_self">tvm::runtime::metadata::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. ">MetadataA [...]
+<tr id="row_113_208_3_" 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="classtvm_1_1runtime_1_1metadata_1_1TensorInfo.html" target="_self">tvm::runtime::metadata::TensorInfo</a></td><td class="desc"></td></tr>
+<tr id="row_113_209_" 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="classtvm_1_1runtime_1_1Module.html" target="_self">tvm::runtime::Module</a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Module.html" title="Module container of TVM. ">Module</a> container of TVM </td></tr>
+<tr id="row_113_210_" 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="classtvm_1_1runtime_1_1NDArray.html" target="_self">tvm::runtime::NDArray</a></td><td class="desc">Managed <a class="el" href="classtvm_1_1runtime_1_1NDArray.html" title="Managed NDArray. The array is backed by reference counted blocks. ">NDArray</a>. The array is backed by referenc [...]
+<tr id="row_113_211_" 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="classtvm_1_1runtime_1_1Optional.html" target="_self">tvm::runtime::Optional< T ></a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Optional.html" title="Optional container that to represent to a Nullable variant of T. ">Optional</a> container that to represen [...]
+<tr id="row_113_212_" 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="classtvm_1_1runtime_1_1PackedFunc.html" target="_self">tvm::runtime::PackedFunc</a></td><td class="desc">Packed function is a type-erased function. The arguments are passed by packed format </td></tr>
+<tr id="row_113_213_" 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="classtvm_1_1runtime_1_1profiling_1_1DeviceWrapper.html" target="_self">tvm::runtime::profiling::DeviceWrapper</a></td><td class="desc">Wrapper for <code>Device</code> </td></tr>
+<tr id="row_113_214_" 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="classtvm_1_1runtime_1_1profiling_1_1MetricCollector.html" target="_self">tvm::runtime::profiling::MetricCollector</a></td><td class="desc">Wrapper for <code><a class="el" href="classtvm_1_1runtime_1_1profiling_1_1MetricCollectorNode.html" title="Interface for user defined profiling [...]
+<tr id="row_113_215_" 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="classtvm_1_1runtime_1_1profiling_1_1Report.html" target="_self">tvm::runtime::profiling::Report</a></td><td class="desc"></td></tr>
+<tr id="row_113_216_" 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="classtvm_1_1runtime_1_1ShapeTuple.html" target="_self">tvm::runtime::ShapeTuple</a></td><td class="desc">Reference to shape tuple objects </td></tr>
+<tr id="row_113_217_" 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="classtvm_1_1runtime_1_1String.html" target="_self">tvm::runtime::String</a></td><td class="desc">Reference to string objects </td></tr>
+<tr id="row_113_218_" 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="classtvm_1_1runtime_1_1Timer.html" target="_self">tvm::runtime::Timer</a></td><td class="desc"><a class="el" href="classtvm_1_1runtime_1_1Timer.html" title="Timer for a specific device. ">Timer</a> for a specific device </td></tr>
+<tr id="row_113_219_" 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="classtvm_1_1runtime_1_1vm_1_1Storage.html" target="_self">tvm::runtime::vm::Storage</a></td><td class="desc">Reference to storage </td></tr>
+<tr id="row_113_220_" 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="classtvm_1_1script_1_1ir__builder_1_1IRBuilder.html" target="_self">tvm::script::ir_builder::IRBuilder</a></td><td class="desc">Managed reference to an <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderNode.html" title="A dialect-agnostic IRBuilder that constructs an [...]
+<tr id="row_113_221_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_221_" class="arrow" onclick="toggleFolder('113_221_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1IRBuilderFrame.html" target="_self">tvm::script::ir_builder::IRBuilderFrame</a></td><td class="desc">Managed reference to an <a class="el" href="classtvm_1_1sc [...]
+<tr id="row_113_221_0_" 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="classtvm_1_1script_1_1ir__builder_1_1ir_1_1IRModuleFrame.html" target="_self">tvm::script::ir_builder::ir::IRModuleFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1ir_1_1IRModuleFrameNode.html" title="A frame that represe [...]
+<tr id="row_113_221_1_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_221_1_" class="arrow" onclick="toggleFolder('113_221_1_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1TIRFrame.html" target="_self">tvm::script::ir_builder::tir::TIRFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_ [...]
+<tr id="row_113_221_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_1script_1_1ir__builder_1_1tir_1_1AllocateConstFrame.html" target="_self">tvm::script::ir_builder::tir::AllocateConstFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AllocateConstFrameNode.html" title=" [...]
+<tr id="row_113_221_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_1script_1_1ir__builder_1_1tir_1_1AllocateFrame.html" target="_self">tvm::script::ir_builder::tir::AllocateFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AllocateFrameNode.html" title="A frame represe [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1AssertFrame.html" target="_self">tvm::script::ir_builder::tir::AssertFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AssertFrameNode.html" title="A frame that represen [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1AttrFrame.html" target="_self">tvm::script::ir_builder::tir::AttrFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1AttrFrameNode.html" title="A frame that represents att [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1BlockFrame.html" target="_self">tvm::script::ir_builder::tir::BlockFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1BlockFrameNode.html" title="A frame that represents [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1BlockInitFrame.html" target="_self">tvm::script::ir_builder::tir::BlockInitFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1BlockInitFrameNode.html" title="A frame that [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1DeclBufferFrame.html" target="_self">tvm::script::ir_builder::tir::DeclBufferFrame</a></td><td class="desc"></td></tr>
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1ElseFrame.html" target="_self">tvm::script::ir_builder::tir::ElseFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ElseFrameNode.html" title="A frame that represents els [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1ForFrame.html" target="_self">tvm::script::ir_builder::tir::ForFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ForFrameNode.html" title="A frame that represents the fo [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1IfFrame.html" target="_self">tvm::script::ir_builder::tir::IfFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1IfFrameNode.html" title="A frame that represents if statem [...]
+<tr id="row_113_221_1_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_1script_1_1ir__builder_1_1tir_1_1LaunchThreadFrame.html" target="_self">tvm::script::ir_builder::tir::LaunchThreadFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1LaunchThreadFrameNode.html" title="Th [...]
+<tr id="row_113_221_1_11_" 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_1script_1_1ir__builder_1_1tir_1_1LetFrame.html" target="_self">tvm::script::ir_builder::tir::LetFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1LetFrameNode.html" title="A frame represents the let bi [...]
+<tr id="row_113_221_1_12_" 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_1script_1_1ir__builder_1_1tir_1_1PrimFuncFrame.html" target="_self">tvm::script::ir_builder::tir::PrimFuncFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1PrimFuncFrameNode.html" title="A frame that r [...]
+<tr id="row_113_221_1_13_" 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_1script_1_1ir__builder_1_1tir_1_1RealizeFrame.html" target="_self">tvm::script::ir_builder::tir::RealizeFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1RealizeFrameNode.html" title="A frame that repr [...]
+<tr id="row_113_221_1_14_" 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_1script_1_1ir__builder_1_1tir_1_1ThenFrame.html" target="_self">tvm::script::ir_builder::tir::ThenFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1ThenFrameNode.html" title="A frame that represents th [...]
+<tr id="row_113_221_1_15_" 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_1script_1_1ir__builder_1_1tir_1_1WhileFrame.html" target="_self">tvm::script::ir_builder::tir::WhileFrame</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1script_1_1ir__builder_1_1tir_1_1WhileFrameNode.html" title="A frame that represents [...]
+<tr id="row_113_222_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_222_" class="arrow" onclick="toggleFolder('113_222_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1Doc.html" target="_self">tvm::script::printer::Doc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1DocNode.html" ti [...]
+<tr id="row_113_222_0_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_222_0_" class="arrow" onclick="toggleFolder('113_222_0_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1ExprDoc.html" target="_self">tvm::script::printer::ExprDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1Ex [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1AttrAccessDoc.html" target="_self">tvm::script::printer::AttrAccessDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AttrAccessDocNode.html" title="Doc that represents attribute access on another [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1CallDoc.html" target="_self">tvm::script::printer::CallDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1CallDocNode.html" title="Doc that represents function call. ">CallDocNode</a> </td></tr>
+<tr id="row_113_222_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_1script_1_1printer_1_1DictDoc.html" target="_self">tvm::script::printer::DictDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1DictDocNode.html" title="Doc that represents dictionary literal. ">DictDocNode</a> </td></tr>
+<tr id="row_113_222_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_1script_1_1printer_1_1IdDoc.html" target="_self">tvm::script::printer::IdDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IdDocNode.html" title="Doc that represents identifier. ">IdDocNode</a> </td></tr>
+<tr id="row_113_222_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_1script_1_1printer_1_1IndexDoc.html" target="_self">tvm::script::printer::IndexDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IndexDocNode.html" title="Doc that represents index access on another expression. ">Index [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1LambdaDoc.html" target="_self">tvm::script::printer::LambdaDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1LambdaDocNode.html" title="Doc that represents anonymous function. ">LambdaDocNode</a> [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1ListDoc.html" target="_self">tvm::script::printer::ListDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ListDocNode.html" title="Doc that represents list literal. ">ListDocNode</a> </td></tr>
+<tr id="row_113_222_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_1script_1_1printer_1_1LiteralDoc.html" target="_self">tvm::script::printer::LiteralDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1LiteralDocNode.html" title="Doc that represents literal value. ">LiteralDocNode</a> < [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1OperationDoc.html" target="_self">tvm::script::printer::OperationDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1OperationDocNode.html" title="Doc that represents operation. ">OperationDocNode</ [...]
+<tr id="row_113_222_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_1script_1_1printer_1_1TupleDoc.html" target="_self">tvm::script::printer::TupleDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1TupleDocNode.html" title="Doc that represents tuple literal. ">TupleDocNode</a> </td></tr>
+<tr id="row_113_222_1_" 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="classtvm_1_1script_1_1printer_1_1SliceDoc.html" target="_self">tvm::script::printer::SliceDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1SliceDocNode.html" title="Doc that represents slice in Index expression. ">SliceDocNode</a> [...]
+<tr id="row_113_222_2_" 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="classtvm_1_1script_1_1printer_1_1StmtBlockDoc.html" target="_self">tvm::script::printer::StmtBlockDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1StmtBlockDocNode.html" title="The container doc that holds a list of StmtDoc. ">Stm [...]
+<tr id="row_113_222_3_" class="even" style="display:none;"><td class="entry"><span style="width:32px;display:inline-block;"> </span><span id="arr_113_222_3_" class="arrow" onclick="toggleFolder('113_222_3_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1script_1_1printer_1_1StmtDoc.html" target="_self">tvm::script::printer::StmtDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1St [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1AssertDoc.html" target="_self">tvm::script::printer::AssertDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AssertDocNode.html" title="Doc that represents assert statement. ">AssertDocNode</a> </ [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1AssignDoc.html" target="_self">tvm::script::printer::AssignDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1AssignDocNode.html" title="Doc that represents assign statement. ">AssignDocNode</a> </ [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1ClassDoc.html" target="_self">tvm::script::printer::ClassDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ClassDocNode.html" title="Doc that represents class definition. ">ClassDocNode</a> </td></tr>
+<tr id="row_113_222_3_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_1script_1_1printer_1_1ExprStmtDoc.html" target="_self">tvm::script::printer::ExprStmtDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ExprStmtDocNode.html" title="Doc that represents an expression as statement. ">Expr [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1ForDoc.html" target="_self">tvm::script::printer::ForDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ForDocNode.html" title="Doc that represents for statement. ">ForDocNode</a> </td></tr>
+<tr id="row_113_222_3_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_1script_1_1printer_1_1FunctionDoc.html" target="_self">tvm::script::printer::FunctionDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1FunctionDocNode.html" title="Doc that represents function definition. ">FunctionDoc [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1IfDoc.html" target="_self">tvm::script::printer::IfDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IfDocNode.html" title="Doc that represent if-then-else statement. ">IfDocNode</a> </td></tr>
+<tr id="row_113_222_3_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_1script_1_1printer_1_1ReturnDoc.html" target="_self">tvm::script::printer::ReturnDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ReturnDocNode.html" title="Doc that represents return statement. ">ReturnDocNode</a> </ [...]
+<tr id="row_113_222_3_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_1script_1_1printer_1_1ScopeDoc.html" target="_self">tvm::script::printer::ScopeDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1ScopeDocNode.html" title="Doc that represents special scopes. ">ScopeDocNode</a> </td></tr>
+<tr id="row_113_222_3_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_1script_1_1printer_1_1WhileDoc.html" target="_self">tvm::script::printer::WhileDoc</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1WhileDocNode.html" title="Doc that represents while statement. ">WhileDocNode</a> </td></tr>
+<tr id="row_113_223_" 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="classtvm_1_1script_1_1printer_1_1Frame.html" target="_self">tvm::script::printer::Frame</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1FrameNode.html">FrameNode</a> </td></tr>
+<tr id="row_113_224_" 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="classtvm_1_1script_1_1printer_1_1IRDocsifier.html" target="_self">tvm::script::printer::IRDocsifier</a></td><td class="desc">Reference type of <a class="el" href="classtvm_1_1script_1_1printer_1_1IRDocsifierNode.html" title="IRDocsifier is the top-level interface in the IR->Doc p [...]
+<tr id="row_113_225_" 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="classtvm_1_1SourceName.html" target="_self">tvm::SourceName</a></td><td class="desc">The source name of a file span </td></tr>
+<tr id="row_113_226_" 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="classtvm_1_1Span.html" target="_self">tvm::Span</a></td><td class="desc"></td></tr>
+<tr id="row_113_227_" 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="classtvm_1_1Target.html" target="_self">tvm::Target</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetNode.html" title="Compilation target. ">TargetNode</a> </td></tr>
+<tr id="row_113_228_" 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="classtvm_1_1TargetKind.html" target="_self">tvm::TargetKind</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetKindNode.html" title="Target kind, specifies the kind of the target. ">TargetKindNode</a> </td></tr>
+<tr id="row_113_229_" 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="classtvm_1_1TargetTag.html" target="_self">tvm::TargetTag</a></td><td class="desc">Managed reference class to <a class="el" href="classtvm_1_1TargetTagNode.html" title="A target tag. ">TargetTagNode</a> </td></tr>
+<tr id="row_113_230_" 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="classtvm_1_1te_1_1IterVarAttr.html" target="_self">tvm::te::IterVarAttr</a></td><td class="desc">Additional scheduable attributes about IterVar </td></tr>
+<tr id="row_113_231_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_231_" class="arrow" onclick="toggleFolder('113_231_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1te_1_1IterVarRelation.html" target="_self">tvm::te::IterVarRelation</a></td><td class="desc">The schedule relation between IterVars can be <a class="el" href="classtvm_1_1te_1_1Split.h [...]
+<tr id="row_113_231_0_" 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="classtvm_1_1te_1_1Fuse.html" target="_self">tvm::te::Fuse</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1FuseNode.html" title="Fuse two domains into one domain. ">FuseNode</a> </td></tr>
+<tr id="row_113_231_1_" 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="classtvm_1_1te_1_1Rebase.html" target="_self">tvm::te::Rebase</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1RebaseNode.html" title="Rebase the iteration to make min to be 0. This is useful to normalize the Schedule to make every leaf...">Rebas [...]
+<tr id="row_113_231_2_" 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="classtvm_1_1te_1_1Singleton.html" target="_self">tvm::te::Singleton</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1SingletonNode.html" title="Singleton iterator [0, 1) ">SingletonNode</a> </td></tr>
+<tr id="row_113_231_3_" 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="classtvm_1_1te_1_1Split.html" target="_self">tvm::te::Split</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1SplitNode.html" title="Split the parent domain into product of outer and iter. ">SplitNode</a> </td></tr>
+<tr id="row_113_231_4_" 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="classtvm_1_1te_1_1Transform.html" target="_self">tvm::te::Transform</a></td><td class="desc"></td></tr>
+<tr id="row_113_232_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_232_" class="arrow" onclick="toggleFolder('113_232_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1te_1_1Operation.html" target="_self">tvm::te::Operation</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Operation.html" title="Operation that produces tensors. ">Operati [...]
+<tr id="row_113_232_0_" 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="classtvm_1_1te_1_1ComputeOp.html" target="_self">tvm::te::ComputeOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ComputeOpNode.html" title="A Compute op that compute a tensor on certain domain. ">ComputeOpNode</a> </td></tr>
+<tr id="row_113_232_1_" 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="classtvm_1_1te_1_1ExternOp.html" target="_self">tvm::te::ExternOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ExternOpNode.html" title="External computation that cannot be splitted. ">ExternOpNode</a> </td></tr>
+<tr id="row_113_232_2_" 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="classtvm_1_1te_1_1HybridOp.html" target="_self">tvm::te::HybridOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1HybridOpNode.html" title="A computation operator that generated by hybrid script. ">HybridOpNode</a> </td></tr>
+<tr id="row_113_232_3_" 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="classtvm_1_1te_1_1PlaceholderOp.html" target="_self">tvm::te::PlaceholderOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1PlaceholderOpNode.html" title="A placeholder op represents an input placeholder. ">PlaceholderOpNode</a> </td></tr>
+<tr id="row_113_232_4_" 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="classtvm_1_1te_1_1ScanOp.html" target="_self">tvm::te::ScanOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1ScanOpNode.html" title="Symbolic scan. ">ScanOpNode</a> </td></tr>
+<tr id="row_113_232_5_" 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="classtvm_1_1te_1_1TensorComputeOp.html" target="_self">tvm::te::TensorComputeOp</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorComputeOpNode.html" title="A TenorCompute op that compute a tensor with an tensor intrinsic. ">TensorComputeOpN [...]
+<tr id="row_113_233_" 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="classtvm_1_1te_1_1Schedule.html" target="_self">tvm::te::Schedule</a></td><td class="desc">Global schedule container For operations and all the operations they depend on. The schedule per <a class="el" href="classtvm_1_1te_1_1Operation.html" title="Operation that produces tensors. " [...]
+<tr id="row_113_234_" 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="classtvm_1_1te_1_1SpecializedCondition.html" target="_self">tvm::te::SpecializedCondition</a></td><td class="desc">Specialized condition to enable op specialization </td></tr>
+<tr id="row_113_235_" 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="classtvm_1_1te_1_1Stage.html" target="_self">tvm::te::Stage</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Stage.html" title="Stage, contains scheduling for a stage of computation. ">Stage</a>, contains scheduling for a stage of computation </td></tr>
+<tr id="row_113_236_" 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="classtvm_1_1te_1_1TensorIntrin.html" target="_self">tvm::te::TensorIntrin</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorIntrinNode.html" title="Node to represent a Tensor intrinsic operator. ">TensorIntrinNode</a> </td></tr>
+<tr id="row_113_237_" 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="classtvm_1_1te_1_1TensorIntrinCall.html" target="_self">tvm::te::TensorIntrinCall</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1te_1_1TensorIntrinCallNode.html">TensorIntrinCallNode</a> </td></tr>
+<tr id="row_113_238_" 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="classtvm_1_1tir_1_1BijectiveLayout.html" target="_self">tvm::tir::BijectiveLayout</a></td><td class="desc">Bijective function mapping for data layout transformation. Given two <a class="el" href="classtvm_1_1tir_1_1Layout.html" title="Managed reference to LayoutNode. ">Layout</a>, < [...]
+<tr id="row_113_239_" 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="classtvm_1_1tir_1_1BlockRV.html" target="_self">tvm::tir::BlockRV</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockRVNode.html" title="A random variable that evaluates to a TensorIR block. ">BlockRVNode</a> </td></tr>
+<tr id="row_113_240_" 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="classtvm_1_1tir_1_1BlockScope.html" target="_self">tvm::tir::BlockScope</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BlockScopeNode.html" title="An object with 1-to-1 correspondence with each block reference in the sref tree. This data structu [...]
+<tr id="row_113_241_" 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="classtvm_1_1tir_1_1Buffer.html" target="_self">tvm::tir::Buffer</a></td><td class="desc"><a class="el" href="classtvm_1_1tir_1_1Buffer.html" title="Buffer is a symbolic n-darray structure. It is a composition of primitive symbolic types...">Buffer</a> is a symbolic n-darray structur [...]
+<tr id="row_113_242_" 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="classtvm_1_1tir_1_1BufferRegion.html" target="_self">tvm::tir::BufferRegion</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1BufferRegionNode.html" title="Representing the region of multi-dimensional buffer access. ">BufferRegionNode</a> </td></tr>
+<tr id="row_113_243_" 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="classtvm_1_1tir_1_1CommReducer.html" target="_self">tvm::tir::CommReducer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1CommReducerNode.html" title="A commutative reducer node to represent a commutative binary operator with identity element..." [...]
+<tr id="row_113_244_" class="even" style="display:none;"><td class="entry"><span style="width:16px;display:inline-block;"> </span><span id="arr_113_244_" class="arrow" onclick="toggleFolder('113_244_')">►</span><span class="icona"><span class="icon">C</span></span><a class="el" href="classtvm_1_1tir_1_1DataProducer.html" target="_self">tvm::tir::DataProducer</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DataProducerNode.html" title="Base [...]
+<tr id="row_113_244_0_" 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="classtvm_1_1te_1_1Tensor.html" target="_self">tvm::te::Tensor</a></td><td class="desc"><a class="el" href="classtvm_1_1te_1_1Tensor.html" title="Tensor structure representing a possible input, or intermediate computation result. ">Tensor</a> structure representing a possible input [...]
+<tr id="row_113_245_" 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="classtvm_1_1tir_1_1Dependency.html" target="_self">tvm::tir::Dependency</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1DependencyNode.html" title="A tuple (src, dst, kind) representing certain types of dependency. For example, (A, B, kRAW) means [...]
+<tr id="row_113_246_" 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="classtvm_1_1tir_1_1IndexMap.html" target="_self">tvm::tir::IndexMap</a></td><td class="desc"></td></tr>
+<tr id="row_113_247_" 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="classtvm_1_1tir_1_1Instruction.html" target="_self">tvm::tir::Instruction</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1InstructionNode.html" title="Schedule instructions each corresponds to a schedule primitive. ">InstructionNode</a> </td></tr>
+<tr id="row_113_248_" 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="classtvm_1_1tir_1_1InstructionKind.html" target="_self">tvm::tir::InstructionKind</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1InstructionKindNode.html" title="Kind of an instruction, e.g. Split, Reorder, etc. Besides the name, every kind of i [...]
+<tr id="row_113_249_" 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="classtvm_1_1tir_1_1IterVar.html" target="_self">tvm::tir::IterVar</a></td><td class="desc">Iteration Variable, represents an iteration over an integer interval </td></tr>
+<tr id="row_113_250_" 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="classtvm_1_1tir_1_1Layout.html" target="_self">tvm::tir::Layout</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LayoutNode.html" title="Layout is to describe how data is organized within an N-dimention tensor. It is composed of upper cas...">Layo [...]
+<tr id="row_113_251_" 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="classtvm_1_1tir_1_1LoopRV.html" target="_self">tvm::tir::LoopRV</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1LoopRVNode.html" title="A random variable that evaluates to a TensorIR for loop. ">LoopRVNode</a> </td></tr>
+<tr id="row_113_252_" 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="classtvm_1_1tir_1_1MatchBufferRegion.html" target="_self">tvm::tir::MatchBufferRegion</a></td><td class="desc">Managed reference to <a class="el" href="classtvm_1_1tir_1_1MatchBufferRegionNode.html" title="Match introduces a constraint that the source buffer region can be remapped t [...]
... 17457 lines suppressed ...