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 2023/01/27 07:34:25 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@18b7dc1dd928bd3e621ff68ad07b354974be13a8)
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 2adc9c38ae deploying docs (apache/tvm@18b7dc1dd928bd3e621ff68ad07b354974be13a8)
2adc9c38ae is described below
commit 2adc9c38ae17659b54b9c51cc2dedc8a6ab5ef61
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Fri Jan 27 07:34:17 2023 +0000
deploying docs (apache/tvm@18b7dc1dd928bd3e621ff68ad07b354974be13a8)
---
docs/_images/sphx_glr_micro_train_001.png | Bin 324292 -> 322817 bytes
docs/_images/sphx_glr_micro_train_thumb.png | Bin 23851 -> 23407 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 | 7 +-
.../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 | 2115 ++++++++++++--------
.../tune_network_cuda.rst.txt | 4 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 39 +-
.../tune_with_autotvm/sg_execution_times.rst.txt | 8 +-
.../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 | 12 +-
.../work_with_relay/sg_execution_times.rst.txt | 8 +-
.../how_to/work_with_schedules/intrin_math.rst.txt | 2 +-
.../work_with_schedules/sg_execution_times.rst.txt | 12 +-
.../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 +-
.../tutorials/optimize/sg_execution_times.rst.txt | 6 +-
.../topic/vta/tutorials/sg_execution_times.rst.txt | 6 +-
.../tutorial/auto_scheduler_matmul_x86.rst.txt | 11 +-
docs/_sources/tutorial/autotvm_matmul_x86.rst.txt | 20 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 56 +-
.../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 | 42 +-
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 | 12 +-
docs/how_to/compile_models/from_pytorch.html | 8 +-
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 | 3 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 40 +-
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 | 39 +-
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 | 2115 ++++++++++++--------
.../tune_with_autoscheduler/tune_network_cuda.html | 4 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 39 +-
.../tune_with_autotvm/sg_execution_times.html | 8 +-
.../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 | 5 +-
docs/how_to/work_with_microtvm/micro_train.html | 16 +-
.../work_with_microtvm/sg_execution_times.html | 12 +-
.../how_to/work_with_relay/sg_execution_times.html | 8 +-
docs/how_to/work_with_schedules/intrin_math.html | 2 +-
.../work_with_schedules/sg_execution_times.html | 12 +-
docs/how_to/work_with_schedules/tensorize.html | 2 +-
docs/reference/api/python/auto_scheduler.html | 4 +-
.../api/typedoc/classes/bytestreamreader.html | 12 +-
.../api/typedoc/classes/cachedcallstack.html | 34 +-
docs/reference/api/typedoc/classes/dldatatype.html | 12 +-
docs/reference/api/typedoc/classes/dldevice.html | 10 +-
.../reference/api/typedoc/classes/environment.html | 12 +-
docs/reference/api/typedoc/classes/ffilibrary.html | 20 +-
.../api/typedoc/classes/graphexecutor.html | 16 +-
docs/reference/api/typedoc/classes/instance.html | 40 +-
docs/reference/api/typedoc/classes/memory.html | 34 +-
docs/reference/api/typedoc/classes/module.html | 10 +-
docs/reference/api/typedoc/classes/ndarray.html | 22 +-
.../api/typedoc/classes/packedfunccell.html | 6 +-
docs/reference/api/typedoc/classes/rpcserver.html | 14 +-
docs/reference/api/typedoc/classes/scalar.html | 6 +-
.../api/typedoc/classes/webgpucontext.html | 12 +-
docs/reference/api/typedoc/enums/argtypecode.html | 30 +-
.../api/typedoc/enums/aynccallbackcode.html | 4 +-
.../api/typedoc/enums/dldatatypecode.html | 8 +-
.../api/typedoc/enums/rpcserverstate.html | 12 +-
docs/reference/api/typedoc/enums/sizeof.html | 18 +-
docs/reference/api/typedoc/index.html | 112 +-
.../api/typedoc/interfaces/disposable.html | 2 +-
.../api/typedoc/interfaces/functioninfo.html | 6 +-
.../api/typedoc/interfaces/libraryprovider.html | 4 +-
docs/searchindex.js | 2 +-
.../vta/tutorials/autotvm/sg_execution_times.html | 4 +-
.../tutorials/frontend/deploy_classification.html | 2 +-
.../vta/tutorials/frontend/deploy_detection.html | 2 +-
.../vta/tutorials/frontend/sg_execution_times.html | 6 +-
.../vta/tutorials/optimize/sg_execution_times.html | 6 +-
docs/topic/vta/tutorials/sg_execution_times.html | 6 +-
docs/tutorial/auto_scheduler_matmul_x86.html | 7 +-
docs/tutorial/autotvm_matmul_x86.html | 20 +-
docs/tutorial/autotvm_relay_x86.html | 268 +--
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 26 +-
docs/tutorial/tensor_expr_get_started.html | 42 +-
129 files changed, 3525 insertions(+), 3285 deletions(-)
diff --git a/docs/_images/sphx_glr_micro_train_001.png b/docs/_images/sphx_glr_micro_train_001.png
index fd04aec899..9d7b73ba75 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 176d23232e..9ae7d8cdc5 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 d5dcb6cbb6..1825363025 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 17.727 seconds)
+ **Total running time of the script:** ( 1 minutes 14.131 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 8dc8ddef08..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 989ms/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 0ddb6773e5..db1ae4d592 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.zipcd4289a4-40e9-4fab-8b7e-a92cf4b3ba76 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip28c6522a-69df-4ac8-ab68-ff22f5a4cbfc 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 1000f767ac..80fac46df8 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, 54.4MB/s]
28%|##7 | 11.5M/41.5M [00:00<00:01, 31.4MB/s]
39%|###8 | 16.0M/41.5M [00:00<00:00, 32.1MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 44.9MB/s]
81%|######## | 33.4M/41.5M [00:00<00:00, 60.0MB/s]
96%|#########6| 39.9M/41.5M [00:00<00:00, 45.3MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 44.1MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
19%|#9 | 7.99M/41.5M [00:00<00:00, 77.5MB/s]
39%|###8 | 16.0M/41.5M [00:00<00:00, 75.0MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 66.7MB/s]
90%|########9 | 37.2M/41.5M [00:00<00:00, 85.3MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 78.9MB/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 baa47ddc3b..8f814998a1 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]
36%|###5 | 16.0M/44.7M [00:00<00:00, 139MB/s]
66%|######5 | 29.3M/44.7M [00:00<00:00, 136MB/s]
95%|#########4| 42.3M/44.7M [00:00<00:00, 90.5MB/s]
100%|##########| 44.7M/44.7M [00:00<00:00, 102MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
27%|##7 | 12.2M/44.7M [00:00<00:00, 128MB/s]
55%|#####4 | 24.4M/44.7M [00:00<00:00, 101MB/s]
80%|#######9 | 35.7M/44.7M [00:00<00:00, 107MB/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 6696e0f291..2699c8edae 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 21.794 seconds)
+ **Total running time of the script:** ( 1 minutes 21.545 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 8a6ad1c747..b6548d8399 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
=================
-**06:24.124** total execution time for **how_to_compile_models** files:
+**06:19.061** total execution time for **how_to_compile_models** files:
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:21.794 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:21.545 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:17.727 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:14.131 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:53.034 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:52.717 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:35.504 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:35.828 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:30.906 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:30.544 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:29.665 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:30.531 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:26.794 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:26.388 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:24.266 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:24.511 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:21.723 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:20.244 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.710 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.622 | 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 391158b044..f820c0874b 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
@@ -727,13 +727,18 @@ 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)
- 2548.1997 2547.6975 2550.0640 2546.8559 1.1696
+ 3341.0008 3340.7200 3343.2723 3338.7640 1.4712
+.. rst-class:: sphx-glr-timing
+
+ **Total running time of the script:** ( 1 minutes 3.163 seconds)
+
+
.. _sphx_glr_download_how_to_deploy_models_deploy_model_on_adreno.py:
.. only:: html
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 46440d1a49..1674480280 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)
- 16.1029 16.0874 16.2248 16.0180 0.0721
+ 16.6069 16.7762 17.1110 16.0513 0.3718
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 ff26f9299d..6470f579c2 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]
5%|4 | 7.99M/170M [00:00<00:02, 74.3MB/s]
9%|9 | 16.0M/170M [00:00<00:02, 71.5MB/s]
14%|#4 | 24.0M/170M [00:00<00:02, 70.8MB/s]
19%|#8 | 32.0M/170M [00:00<00:01, 73.1MB/s]
24%|##3 | 40.0M/170M [00:00<00:01, 73.0MB/s]
28%|##8 | 48.0M/170M [00:00<00:01, 73.8MB/s]
32%|###2 | 55.0M/170M [00:00<00:01, 69.8MB/s]
37%|###6 | 62.3M/170M [00:00<00:01, 62.1MB/s]
42%|####2 | 72.0M/170M [00:01<00:01, 62.6MB/s]
47%|####7 | 80.0M/170M [00:01<00:01, 67.1MB/s]
52%|#####1 | 88.0M/170M [00:01<00:01, 69.5MB/s]
58%|#####8 | 98.7M/170M [00:01<00:00, 80.7MB/s]
63%|######2 | 107M/170M [00:01<00:00, 81.0MB/s]
69%|######9 | 118M/170M [00:01<00:00, 90.3MB/s]
74%|#######4 | 126M/170M [00:01<00:00, 79.8MB/s]
79%|#######9 | 134M/170M [00:01<00:00, 73.5MB/s]
83%|########3 | 142M/170M [00:02<00:00, 71.3MB/s]
88%|########8 | 150M/170M [00:02<00:00, 74.6MB/s]
93%|#########2| 158M/170M [00:02<00:00, 54.4MB/s]
98%|#########7| 166M/170M [00:02<00:00, 62.3MB/s]
100%|##########| 170M/170M [00:02<00:00, 67.3MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
5%|4 | 7.99M/170M [00:00<00:02, 83.3MB/s]
13%|#2 | 21.3M/170M [00:00<00:01, 116MB/s]
19%|#9 | 32.3M/170M [00:00<00:01, 106MB/s]
25%|##5 | 42.6M/170M [00:00<00:01, 105MB/s]
31%|### | 52.6M/170M [00:00<00:01, 103MB/s]
37%|###6 | 62.5M/170M [00:00<00:01, 102MB/s]
43%|####2 | 72.3M/170M [00:00<00:01, 101MB/s]
48%|####8 | 81.9M/170M [00:00<00:00, 96.7MB/s]
55%|#####4 | 92.9M/170M [00:00<00:00, 102MB/s]
60%|###### | 103M/170M [00:01<00:00, 99.3MB/s]
66%|######6 | 112M/170M [00:01<00:00, 100MB/s]
72%|#######2 | 123M/170M [00:01<00:00, 102MB/s]
78%|#######7 | 132M/170M [00:01<00:00, 96.5MB/s]
84%|########4 | 143M/170M [00:01<00:00, 102MB/s]
90%|######### | 153M/170M [00:01<00:00, 88.6MB/s]
99%|#########8| 167M/170M [00:01<00:00, 105MB/s]
100%|##########| 170M/170M [00:01<00:00, 101MB/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 34.456 seconds)
+ **Total running time of the script:** ( 3 minutes 27.786 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 f0a402ee09..c9e5573f32 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, 38.7MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 58.5MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
100%|#########9| 13.5M/13.6M [00:00<00:00, 142MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 141MB/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)
- 90.3979 90.2656 93.2883 90.1057 0.4102
+ 90.2892 90.2321 90.9815 90.0854 0.1888
@@ -458,7 +458,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 15.492 seconds)
+ **Total running time of the script:** ( 1 minutes 14.094 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 3f45e6dd59..fe538b692a 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)
- 120.9549 120.9322 124.1480 120.2381 0.5169
+ 120.8368 120.7806 123.1834 120.0746 0.4294
@@ -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 29.988 seconds)
+ **Total running time of the script:** ( 2 minutes 34.550 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 77f0a649b2..fdc04a776d 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 41.583 seconds)
+ **Total running time of the script:** ( 1 minutes 20.755 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 ee289926a8..28197d4ecd 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]
3%|2 | 3476/132723 [00:00<00:03, 34756.72KB/s]
8%|8 | 10875/132723 [00:00<00:02, 57827.60KB/s]
13%|#2 | 16658/132723 [00:00<00:04, 27190.02KB/s]
16%|#6 | 21691/132723 [00:00<00:03, 30515.28KB/s]
23%|##2 | 29999/132723 [00:00<00:02, 43065.12KB/s]
27%|##6 | 35488/132723 [00:00<00:02, 40446.62KB/s]
30%|### | 40317/132723 [00:01<00:02, 40208.61KB/s]
37%|###6 | 48976/132723 [00:01<00:01, 51664.61KB/s]
41%|####1 | 54831/132723 [00:01<00:01, 46908.56KB/s]
46%|####5 | 60726/132723 [00:01<00:01, 49868.54KB/s]
50%|####9 | 66151/132723 [00:01<00:01, 49220.32KB/s]
56%|#####6 | 74670/132723 [00:01<00:00, 58717.54KB/s]
62%|######1 | 81905/132723 [00:01<00:00, 55924.82KB/s]
68%|######8 | 90426/132723 [00:01<00:00, 63522.73KB/s]
74%|#######4 | 98288/132723 [00:01<00:00, 62993.22KB/s]
80%|######## |
106665/132723 [00:02<00:00, 68490.96KB/s]
87%|########6 | 115234/132723 [00:02<00:00, 73208.73KB/s]
93%|#########3| 123883/132723 [00:02<00:00, 76946.13KB/s]
100%|#########9| 132591/132723 [00:02<00:00, 79852.22KB/s]
100%|##########| 132723/132723 [00:02<00:00, 55436.23KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
5%|4 | 6147/132723 [00:00<00:02, 61464.05KB/s]
11%|# | 14511/132723 [00:00<00:01, 74505.85KB/s]
17%|#6 | 21962/132723 [00:00<00:01, 59073.28KB/s]
21%|##1 | 28175/132723 [00:00<00:01, 59562.02KB/s]
27%|##6 | 35273/132723 [00:00<00:01, 63275.70KB/s]
33%|###3 | 44041/132723 [00:00<00:01, 71017.19KB/s]
40%|###9 | 52694/132723 [00:00<00:01, 75854.81KB/s]
46%|####6 | 61385/132723 [00:00<00:00, 79259.51KB/s]
53%|#####2 | 70096/132723 [00:00<00:00, 81651.16KB/s]
59%|#####9 | 78933/132723 [00:01<00:00, 83691.51KB/s]
66%|######5 | 87584/132723 [00:01<00:00, 84539.97KB/s]
73%|#######2 | 96432/132723 [00:01<00:00, 85722.14KB/s]
79%|#######9 | 105229/132723 [00:01<00:00, 86395.14KB/s]
86%|########5 | 114055/132723 [00:01<00:00, 86954.14KB/s]
93%|#########2| 122821/132723 [00:01<00:00, 87162.81KB/s]
99%|########
#9| 131635/132723 [00:01<00:00, 87454.03KB/s]
100%|##########| 132723/132723 [00:01<00:00, 79248.71KB/s]
@@ -246,7 +246,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 3 minutes 40.281 seconds)
+ **Total running time of the script:** ( 3 minutes 34.842 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 c462be5fe4..107623b996 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
=================
-**15:12.314** total execution time for **how_to_deploy_models** files:
+**14:51.380** total execution time for **how_to_deploy_models** files:
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:40.281 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 03:34.842 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:34.456 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:27.786 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:29.988 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:34.550 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:41.583 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:20.755 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:15.492 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:14.094 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:54.292 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 01:03.163 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:41.392 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:40.880 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:27.598 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:27.715 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:27.226 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:27.588 | 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 51d39fe1f0..39a4c7aa80 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.zip2840830e-0a34-445f-b9f6-dc7e4a1cd4ed from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip2e21da86-954a-4776-ad44-0a8bcc9cb300 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 748c3e3090..57719ef473 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
=================
-**00:53.794** total execution time for **how_to_extend_tvm** files:
+**00:53.112** 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``) | 00:49.939 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:49.323 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.758 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.708 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.090 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.074 | 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 9b2bf3f463..81a9aa1a59 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: 21272us [21272us] (48.68%; 48.68%)
- FoldScaleAxis: 22428us [7us] (51.32%; 51.32%)
- FoldConstant: 22420us [1736us] (51.30%; 99.97%)
- InferType: 20684us [20684us] (47.33%; 92.26%)
+ InferType: 22115us [22115us] (49.77%; 49.77%)
+ FoldScaleAxis: 22321us [8us] (50.23%; 50.23%)
+ FoldConstant: 22313us [1754us] (50.21%; 99.97%)
+ InferType: 20559us [20559us] (46.27%; 92.14%)
@@ -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: 20703us [20703us] (48.25%; 48.25%)
- FoldScaleAxis: 22200us [5us] (51.75%; 51.75%)
- FoldConstant: 22195us [1713us] (51.73%; 99.98%)
- InferType: 20482us [20482us] (47.74%; 92.28%)
+ InferType: 20728us [20728us] (48.25%; 48.25%)
+ FoldScaleAxis: 22232us [6us] (51.75%; 51.75%)
+ FoldConstant: 22227us [1738us] (51.74%; 99.98%)
+ InferType: 20489us [20489us] (47.69%; 92.18%)
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 69b0b200ff..c232f2348b 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: 54.141120 ms
+ Convolution: 36.795871 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 a389c24e87..c81df42083 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
@@ -608,7 +608,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 7.909705 ms
+ conv2d with tensor core: 7.718617 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 5de9367185..74ee4986ee 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: 0.019201
- Baseline: 3.457197
+ Numpy running time: 0.018811
+ Baseline: 3.396133
@@ -227,7 +227,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.305233
+ Opt1: 0.302338
@@ -318,7 +318,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.338993
+ Opt2: 0.332977
@@ -406,7 +406,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.118452
+ Opt3: 0.116724
@@ -523,7 +523,7 @@ flattening.
.. code-block:: none
- Opt4: 0.109763
+ Opt4: 0.109741
@@ -635,7 +635,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.111094
+ Opt5: 0.110908
@@ -748,7 +748,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
.. code-block:: none
- Opt6: 0.148265
+ Opt6: 0.146963
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 9c835637a5..5e4eac07cc 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
=================
-**00:35.477** total execution time for **how_to_optimize_operators** files:
+**00:34.845** total execution time for **how_to_optimize_operators** files:
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.722 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.301 | 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_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.475 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.203 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.068 | 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 03c0d3cbb8..4e3b22c967 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
=================
-**09:23.380** total execution time for **how_to_tune_with_autoscheduler** files:
+**09:23.537** 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``) | 05:38.539 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:42.565 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:40.846 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:39.693 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:06.966 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:05.839 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:29.196 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:28.284 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.591 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.117 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.241 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.038 | 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 098c70551f..e0972c5114 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
@@ -244,13 +244,13 @@ cooperative fetching, unrolling and operator fusion.
def main(data: T.Buffer((1, 512, 7, 7), "float32"), kernel: T.Buffer((512, 512, 3, 3), "float32"), bias: T.Buffer((1, 512, 1, 1), "float32"), compute: T.Buffer((1, 512, 7, 7), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
blockIdx_x = T.env_thread("blockIdx.x")
- T.launch_thread(blockIdx_x, 8)
- conv2d_nchw = T.allocate([28], "float32", "local")
- pad_temp_shared = T.allocate([1296], "float32", "shared")
- kernel_shared = T.allocate([9216], "float32", "shared")
+ T.launch_thread(blockIdx_x, 112)
+ conv2d_nchw = T.allocate([7], "float32", "local")
+ pad_temp_shared = T.allocate([432], "float32", "shared")
+ kernel_shared = T.allocate([4608], "float32", "shared")
threadIdx_x = T.env_thread("threadIdx.x")
- T.launch_thread(threadIdx_x, 112)
- conv2d_nchw_1 = T.Buffer((28,), data=conv2d_nchw, scope="local")
+ T.launch_thread(threadIdx_x, 32)
+ conv2d_nchw_1 = T.Buffer((7,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
conv2d_nchw_1[1] = T.float32(0)
conv2d_nchw_1[2] = T.float32(0)
@@ -258,457 +258,786 @@ cooperative fetching, unrolling and operator fusion.
conv2d_nchw_1[4] = T.float32(0)
conv2d_nchw_1[5] = T.float32(0)
conv2d_nchw_1[6] = T.float32(0)
- conv2d_nchw_1[7] = T.float32(0)
- conv2d_nchw_1[8] = T.float32(0)
- conv2d_nchw_1[9] = T.float32(0)
- conv2d_nchw_1[10] = T.float32(0)
- conv2d_nchw_1[11] = T.float32(0)
- conv2d_nchw_1[12] = T.float32(0)
- conv2d_nchw_1[13] = T.float32(0)
- conv2d_nchw_1[14] = T.float32(0)
- conv2d_nchw_1[15] = T.float32(0)
- conv2d_nchw_1[16] = T.float32(0)
- conv2d_nchw_1[17] = T.float32(0)
- conv2d_nchw_1[18] = T.float32(0)
- conv2d_nchw_1[19] = T.float32(0)
- conv2d_nchw_1[20] = T.float32(0)
- conv2d_nchw_1[21] = T.float32(0)
- conv2d_nchw_1[22] = T.float32(0)
- conv2d_nchw_1[23] = T.float32(0)
- conv2d_nchw_1[24] = T.float32(0)
- conv2d_nchw_1[25] = T.float32(0)
- conv2d_nchw_1[26] = T.float32(0)
- conv2d_nchw_1[27] = T.float32(0)
for rc_outer_outer in range(32):
cse_var_2: T.int32 = rc_outer_outer * 784
cse_var_1: T.int32 = rc_outer_outer * 144
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((1296,), data=pad_temp_shared, scope="shared")
+ pad_temp_shared_1 = T.Buffer((432,), data=pad_temp_shared, scope="shared")
data_1 = T.Buffer((25088,), data=data.data)
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(9 <= threadIdx_x_1 % 81 and threadIdx_x_1 % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 81 * 49 + threadIdx_x_1 % 81 // 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(9 <= (threadIdx_x_1 + 31) % 81 and (threadIdx_x_1 + 31) % 81 < 72 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 112) // 81 * 49 + (threadIdx_x_1 + 31) % 81 // 9 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(9 <= (threadIdx_x_1 + 62) % 81 and (threadIdx_x_1 + 62) % 81 < 72 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 81 * 49 + (threadIdx_x_1 + 62) % 81 // 9 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(9 <= (threadIdx_x_1 + 12) % 81 and (threadIdx_x_1 + 12) % 81 < 72 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 336) // 81 * 49 + (threadIdx_x_1 + 12) % 81 // 9 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(9 <= (threadIdx_x_1 + 43) % 81 and (threadIdx_x_1 + 43) % 81 < 72 and 1 <= (threadIdx_x_1 + 7) % 9 and (threadIdx_x_1 + 7) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 448) // 81 * 49 + (threadIdx_x_1 + 43) % 81 // 9 * 7 + (threadIdx_x_1 + 7) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(9 <= (threadIdx_x_1 + 74) % 81 and (threadIdx_x_1 + 74) % 81 < 72 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 560) // 81 * 49 + (threadIdx_x_1 + 74) % 81 // 9 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(9 <= (threadIdx_x_1 + 24) % 81 and (threadIdx_x_1 + 24) % 81 < 72 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 672) // 81 * 49 + (threadIdx_x_1 + 24) % 81 // 9 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(9 <= (threadIdx_x_1 + 55) % 81 and (threadIdx_x_1 + 55) % 81 < 72 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 784) // 81 * 49 + (threadIdx_x_1 + 55) % 81 // 9 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(9 <= (threadIdx_x_1 + 5) % 81 and (threadIdx_x_1 + 5) % 81 < 72 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 896) // 81 * 49 + (threadIdx_x_1 + 5) % 81 // 9 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(1 <= (threadIdx_x_1 // 9 + 4) % 9 and (threadIdx_x_1 + 36) % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1008) // 81 * 49 + (threadIdx_x_1 // 9 + 4) % 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(9 <= (threadIdx_x_1 + 67) % 81 and (threadIdx_x_1 + 67) % 81 < 72 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1120) // 81 * 49 + (threadIdx_x_1 + 67) % 81 // 9 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- if T.likely(threadIdx_x_1 < 64):
- pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(threadIdx_x_1 < 55 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1232) // 81 * 49 + (threadIdx_x_1 + 17) % 81 // 9 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 and threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 27 * 49 + threadIdx_x_1 % 27 // 9 * 7 + blockIdx_x % 7 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 32] = T.if_then_else(1 <= (threadIdx_x_1 + 5) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 5) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 32) // 27 * 49 + (threadIdx_x_1 + 5) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 64] = T.if_then_else(1 <= (threadIdx_x_1 + 10) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 10) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 64) // 27 * 49 + (threadIdx_x_1 + 10) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 96] = T.if_then_else(1 <= (threadIdx_x_1 + 15) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 15) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 96) // 27 * 49 + (threadIdx_x_1 + 15) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 128] = T.if_then_else(1 <= (threadIdx_x_1 + 20) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 20) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 128) // 27 * 49 + (threadIdx_x_1 + 20) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 160] = T.if_then_else(1 <= (threadIdx_x_1 + 25) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 25) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 7) % 9 and (threadIdx_x_1 + 7) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 160) // 27 * 49 + (threadIdx_x_1 + 25) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 7) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 192] = T.if_then_else(1 <= (threadIdx_x_1 + 3) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 3) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 192) // 27 * 49 + (threadIdx_x_1 + 3) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= (threadIdx_x_1 + 8) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 8) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 27 * 49 + (threadIdx_x_1 + 8) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 256] = T.if_then_else(1 <= (threadIdx_x_1 + 13) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 13) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 256) // 27 * 49 + (threadIdx_x_1 + 13) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 288] = T.if_then_else(1 <= blockIdx_x % 7 + (threadIdx_x_1 // 9 + 2) % 3 and blockIdx_x % 7 + (threadIdx_x_1 // 9 + 2) % 3 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 288) // 27 * 49 + blockIdx_x % 7 * 7 + (threadIdx_x_1 // 9 + 2) % 3 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 320] = T.if_then_else(1 <= (threadIdx_x_1 + 23) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 23) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 320) // 27 * 49 + (threadIdx_x_1 + 23) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 352] = T.if_then_else(1 <= (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 352) // 27 * 49 + (threadIdx_x_1 + 1) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 384] = T.if_then_else(1 <= (threadIdx_x_1 + 6) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 6) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 384) // 27 * 49 + (threadIdx_x_1 + 6) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ if T.likely(threadIdx_x_1 < 16):
+ pad_temp_shared_1[threadIdx_x_1 + 416] = T.if_then_else((threadIdx_x_1 + 11) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 416) // 27 * 49 + (threadIdx_x_1 + 11) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
- kernel_shared_1 = T.Buffer((9216,), data=kernel_shared, scope="shared")
+ kernel_shared_1 = T.Buffer((4608,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 112) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 336) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 560) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 672) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 784) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 896) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 32256]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1120) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1232) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1344) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1456) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1568) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1680) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1792) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1904) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 64512]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2240) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2352] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2464) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2576] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2576) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2688) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2800] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2800) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2912) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3024] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 96768]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3136] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3136) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3248] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3248) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3360] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3360) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3472] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3472) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3584] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3584) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3696] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3696) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3808] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3808) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3920] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3920) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4032] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 129024]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4144] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4144) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4256] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4368] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4368) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4480] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4592] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4592) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4704] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4704) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4816] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4816) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4928] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4928) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5040] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 161280]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5152] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5152) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5264] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5264) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5376] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5376) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5488] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5488) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5600] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5600) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5712] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5712) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5824] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5824) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5936] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5936) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6048] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 193536]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6160] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6272] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6272) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6384] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6496] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6496) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6608] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6608) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6720] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6720) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6832] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6832) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6944] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6944) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7056] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 225792]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7168] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7168) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7280] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7280) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7392] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7392) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7504] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7504) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7616] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7616) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7728] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7728) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7840] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7840) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7952] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7952) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8064] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 258048]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8176] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8176) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8288] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8288) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8400] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8400) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8512] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8624] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8624) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8736] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8736) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8848] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8848) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8960] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8960) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 9072] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 290304]
- with T.launch_thread(threadIdx_x_2, 112):
- if T.likely(threadIdx_x_2 < 32):
- kernel_shared_1[threadIdx_x_2 + 9184] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 9184) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- for rc_outer_inner, ry_outer_inner, rx_outer_inner in T.grid(2, 3, 3):
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- for i1_inner, i3_inner in T.grid(4, 7):
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 32] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + (threadIdx_x_2 + 32) // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 64] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + (threadIdx_x_2 + 64) // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 96] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 96]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 128] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 160] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 192] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 192) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 256] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 288] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 9216]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 320] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 320) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 352] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 384] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 416] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 416) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 480] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 512] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 544] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 544) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 576] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 18432]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 608] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 608) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 640] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 640) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 672) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 704] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 704) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 736] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 736) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 768] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 768) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 800] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 800) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 832] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 832) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 864] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 27648]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 896) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 928] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 928) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 960] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 960) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 992] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 992) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1024] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1024) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1056] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1056) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1088] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1088) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1120) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1152] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 36864]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1184] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1184) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1216] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1216) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1248] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1248) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1280] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1280) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1312] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1312) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1344) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1376] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1376) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1408] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1408) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1440] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 46080]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1472] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1472) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1504] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1504) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1536] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1536) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1568) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1600] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1600) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1632] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1632) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1664] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1664) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1696] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1696) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1728] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 55296]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1760] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1760) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1792) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1824] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1824) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1856] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1856) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1888] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1888) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1920] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1920) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1952] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1952) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1984] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1984) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 64512]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2048] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2048) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2080] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2080) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2112] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2112) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2144] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2144) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2176] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2176) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2208] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2208) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2240) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2272] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2272) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2304] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 73728]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2336] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2336) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2368] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2368) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2400] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2400) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2432] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2432) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2464) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2496] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2496) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2528] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2528) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2560] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2560) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2592] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 82944]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2624] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2624) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2656] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2656) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2688) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2720] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2720) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2752] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2752) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2784] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2784) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2816] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2816) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2848] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2848) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2880] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 92160]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2912) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2944] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2944) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2976] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2976) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3008] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3008) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3040] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3040) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3072] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3072) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3104] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3104) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3136] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3136) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3168] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 101376]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3200] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3200) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3232] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3232) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3264] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3264) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3296] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3296) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3328] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3328) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3360] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3360) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3392] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3392) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3424] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3424) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3456] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 110592]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3488] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3488) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3520] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3520) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3552] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3552) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3584] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3584) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3616] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3616) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3648] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3648) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3680] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3680) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3712] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3712) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3744] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 119808]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3776] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3776) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3808] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3808) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3840] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3840) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3872] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3872) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3904] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3904) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3936] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3936) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3968] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3968) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4000] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4000) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4032] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 129024]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4064] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4064) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4096] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4096) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4128] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4160] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4192] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4192) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4224] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4256] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4288] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4288) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4320] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 138240]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4352] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4384] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4416] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4416) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4448] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4480] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4512] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4544] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4544) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4576] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4576) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ for ry_outer_inner in range(3):
+ cse_var_115: T.int32 = ry_outer_inner * 9
+ cse_var_114: T.int32 = cse_var_115 + 88
+ cse_var_113: T.int32 = cse_var_115 + 87
+ cse_var_112: T.int32 = cse_var_115 + 86
+ cse_var_111: T.int32 = cse_var_115 + 85
+ cse_var_110: T.int32 = cse_var_115 + 84
+ cse_var_109: T.int32 = cse_var_115 + 83
+ cse_var_108: T.int32 = cse_var_115 + 82
+ cse_var_107: T.int32 = cse_var_115 + 7
+ cse_var_106: T.int32 = cse_var_115 + 61
+ cse_var_105: T.int32 = cse_var_115 + 60
+ cse_var_104: T.int32 = cse_var_115 + 6
+ cse_var_103: T.int32 = cse_var_115 + 59
+ cse_var_102: T.int32 = cse_var_115 + 58
+ cse_var_101: T.int32 = cse_var_115 + 57
+ cse_var_100: T.int32 = cse_var_115 + 56
+ cse_var_99: T.int32 = cse_var_115 + 55
+ cse_var_98: T.int32 = cse_var_115 + 5
+ cse_var_97: T.int32 = cse_var_115 + 412
+ cse_var_96: T.int32 = cse_var_115 + 411
+ cse_var_95: T.int32 = cse_var_115 + 410
+ cse_var_94: T.int32 = cse_var_115 + 409
+ cse_var_93: T.int32 = cse_var_115 + 408
+ cse_var_92: T.int32 = cse_var_115 + 407
+ cse_var_91: T.int32 = cse_var_115 + 406
+ cse_var_90: T.int32 = cse_var_115 + 4
+ cse_var_89: T.int32 = cse_var_115 + 385
+ cse_var_88: T.int32 = cse_var_115 + 384
+ cse_var_87: T.int32 = cse_var_115 + 383
+ cse_var_86: T.int32 = cse_var_115 + 382
+ cse_var_85: T.int32 = cse_var_115 + 381
+ cse_var_84: T.int32 = cse_var_115 + 380
+ cse_var_83: T.int32 = cse_var_115 + 379
+ cse_var_82: T.int32 = cse_var_115 + 358
+ cse_var_81: T.int32 = cse_var_115 + 357
+ cse_var_80: T.int32 = cse_var_115 + 356
+ cse_var_79: T.int32 = cse_var_115 + 355
+ cse_var_78: T.int32 = cse_var_115 + 354
+ cse_var_77: T.int32 = cse_var_115 + 353
+ cse_var_76: T.int32 = cse_var_115 + 352
+ cse_var_75: T.int32 = cse_var_115 + 34
+ cse_var_74: T.int32 = cse_var_115 + 331
+ cse_var_73: T.int32 = cse_var_115 + 330
+ cse_var_72: T.int32 = cse_var_115 + 33
+ cse_var_71: T.int32 = cse_var_115 + 329
+ cse_var_70: T.int32 = cse_var_115 + 328
+ cse_var_69: T.int32 = cse_var_115 + 327
+ cse_var_68: T.int32 = cse_var_115 + 326
+ cse_var_67: T.int32 = cse_var_115 + 325
+ cse_var_66: T.int32 = cse_var_115 + 32
+ cse_var_65: T.int32 = cse_var_115 + 31
+ cse_var_64: T.int32 = cse_var_115 + 304
+ cse_var_63: T.int32 = cse_var_115 + 303
+ cse_var_62: T.int32 = cse_var_115 + 302
+ cse_var_61: T.int32 = cse_var_115 + 301
+ cse_var_60: T.int32 = cse_var_115 + 300
+ cse_var_59: T.int32 = cse_var_115 + 30
+ cse_var_58: T.int32 = cse_var_115 + 3
+ cse_var_57: T.int32 = cse_var_115 + 299
+ cse_var_56: T.int32 = cse_var_115 + 298
+ cse_var_55: T.int32 = cse_var_115 + 29
+ cse_var_54: T.int32 = cse_var_115 + 28
+ cse_var_53: T.int32 = cse_var_115 + 277
+ cse_var_52: T.int32 = cse_var_115 + 276
+ cse_var_51: T.int32 = cse_var_115 + 275
+ cse_var_50: T.int32 = cse_var_115 + 274
+ cse_var_49: T.int32 = cse_var_115 + 273
+ cse_var_48: T.int32 = cse_var_115 + 272
+ cse_var_47: T.int32 = cse_var_115 + 271
+ cse_var_46: T.int32 = cse_var_115 + 250
+ cse_var_45: T.int32 = cse_var_115 + 249
+ cse_var_44: T.int32 = cse_var_115 + 248
+ cse_var_43: T.int32 = cse_var_115 + 247
+ cse_var_42: T.int32 = cse_var_115 + 246
+ cse_var_41: T.int32 = cse_var_115 + 245
+ cse_var_40: T.int32 = cse_var_115 + 244
+ cse_var_39: T.int32 = cse_var_115 + 223
+ cse_var_38: T.int32 = cse_var_115 + 222
+ cse_var_37: T.int32 = cse_var_115 + 221
+ cse_var_36: T.int32 = cse_var_115 + 220
+ cse_var_35: T.int32 = cse_var_115 + 219
+ cse_var_34: T.int32 = cse_var_115 + 218
+ cse_var_33: T.int32 = cse_var_115 + 217
+ cse_var_32: T.int32 = cse_var_115 + 2
+ cse_var_31: T.int32 = cse_var_115 + 196
+ cse_var_30: T.int32 = cse_var_115 + 195
+ cse_var_29: T.int32 = cse_var_115 + 194
+ cse_var_28: T.int32 = cse_var_115 + 193
+ cse_var_27: T.int32 = cse_var_115 + 192
+ cse_var_26: T.int32 = cse_var_115 + 191
+ cse_var_25: T.int32 = cse_var_115 + 190
+ cse_var_24: T.int32 = cse_var_115 + 169
+ cse_var_23: T.int32 = cse_var_115 + 168
+ cse_var_22: T.int32 = cse_var_115 + 167
+ cse_var_21: T.int32 = cse_var_115 + 166
+ cse_var_20: T.int32 = cse_var_115 + 165
+ cse_var_19: T.int32 = cse_var_115 + 164
+ cse_var_18: T.int32 = cse_var_115 + 163
+ cse_var_17: T.int32 = cse_var_115 + 142
+ cse_var_16: T.int32 = cse_var_115 + 141
+ cse_var_15: T.int32 = cse_var_115 + 140
+ cse_var_14: T.int32 = cse_var_115 + 139
+ cse_var_13: T.int32 = cse_var_115 + 138
+ cse_var_12: T.int32 = cse_var_115 + 137
+ cse_var_11: T.int32 = cse_var_115 + 136
+ cse_var_10: T.int32 = cse_var_115 + 115
+ cse_var_9: T.int32 = cse_var_115 + 114
+ cse_var_8: T.int32 = cse_var_115 + 113
+ cse_var_7: T.int32 = cse_var_115 + 112
+ cse_var_6: T.int32 = cse_var_115 + 111
+ cse_var_5: T.int32 = cse_var_115 + 110
+ cse_var_4: T.int32 = cse_var_115 + 109
+ cse_var_3: T.int32 = cse_var_115 + 1
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_3] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_99] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_4] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 135] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_11] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 162] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_18] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 189] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_25] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 216] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_33] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 243] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_40] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 270] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_47] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 297] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_56] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 324] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_67] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 351] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_76] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 378] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_83] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 405] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_91] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_3] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_99] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_4] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_11] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_18] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_25] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_33] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_40] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_47] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_56] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_67] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_76] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_83] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_91] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_107] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_75] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_106] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_114] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_10] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_17] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_24] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_31] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_39] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_46] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_53] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_64] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_74] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_82] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_97] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_107] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_75] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_106] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_114] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_10] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 116] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_17] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 143] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_24] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 170] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_31] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 197] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_39] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 224] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_46] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 251] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_53] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 278] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_64] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 305] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_74] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 332] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_82] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 359] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 386] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_97] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 413] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ for i3_inner in range(7):
compute_1 = T.Buffer((25088,), data=compute.data)
bias_1 = T.Buffer((512,), data=bias.data)
- compute_1[blockIdx_x * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + threadIdx_x % 7 * 7 + i3_inner] = T.max(conv2d_nchw_1[i1_inner * 7 + i3_inner] + bias_1[blockIdx_x * 64 + threadIdx_x // 7 * 4 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 1568 + threadIdx_x * 49 + blockIdx_x % 7 * 7 + i3_inner] = T.max(conv2d_nchw_1[i3_inner] + bias_1[blockIdx_x // 7 * 32 + threadIdx_x], T.float32(0))
@@ -758,7 +1087,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.398 ms
+ Execution time of this operator: 0.251 ms
@@ -806,33 +1135,33 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
- conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=4)
+ 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=16)
+ conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=32)
conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
- conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+ conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
- conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=7)
- 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_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=7)
conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
- conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
- conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=2)
+ conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=16)
+ conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=3)
- conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
- conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+ conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
+ conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2 [...]
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
- compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
- compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=16)
+ 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=32)
compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
- compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+ compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
@@ -855,12 +1184,12 @@ 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=112)
+ kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=32)
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)
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=112)
+ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=32)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
@@ -880,10 +1209,10 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
#define int64_t long long
#define uint64_t unsigned long long
#endif
- extern "C" __global__ void __launch_bounds__(112) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[28];
- __shared__ float pad_temp_shared[1296];
- __shared__ float kernel_shared[9216];
+ extern "C" __global__ void __launch_bounds__(32) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[7];
+ __shared__ float pad_temp_shared[432];
+ __shared__ float kernel_shared[4608];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
conv2d_nchw[2] = 0.000000e+00f;
@@ -891,364 +1220,510 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw[4] = 0.000000e+00f;
conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[6] = 0.000000e+00f;
- conv2d_nchw[7] = 0.000000e+00f;
- conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[9] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[11] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
- conv2d_nchw[13] = 0.000000e+00f;
- conv2d_nchw[14] = 0.000000e+00f;
- conv2d_nchw[15] = 0.000000e+00f;
- conv2d_nchw[16] = 0.000000e+00f;
- conv2d_nchw[17] = 0.000000e+00f;
- conv2d_nchw[18] = 0.000000e+00f;
- conv2d_nchw[19] = 0.000000e+00f;
- conv2d_nchw[20] = 0.000000e+00f;
- conv2d_nchw[21] = 0.000000e+00f;
- conv2d_nchw[22] = 0.000000e+00f;
- conv2d_nchw[23] = 0.000000e+00f;
- conv2d_nchw[24] = 0.000000e+00f;
- conv2d_nchw[25] = 0.000000e+00f;
- conv2d_nchw[26] = 0.000000e+00f;
- conv2d_nchw[27] = 0.000000e+00f;
for (int rc_outer_outer = 0; rc_outer_outer < 32; ++rc_outer_outer) {
__syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = (((((9 <= (((int)threadIdx.x) % 81)) && ((((int)threadIdx.x) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + ((((int)threadIdx.x) / 81) * 49)) + (((((int)threadIdx.x) % 81) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((9 <= ((((int)threadIdx.x) + 31) % 81)) && (((((int)threadIdx.x) + 31) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 112) / 81) * 49)) + ((((((int)threadIdx.x) + 31) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((9 <= ((((int)threadIdx.x) + 62) % 81)) && (((((int)threadIdx.x) + 62) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 224) / 81) * 49)) + ((((((int)threadIdx.x) + 62) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((9 <= ((((int)threadIdx.x) + 12) % 81)) && (((((int)threadIdx.x) + 12) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 336) / 81) * 49)) + ((((((int)threadIdx.x) + 12) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 448)] = (((((9 <= ((((int)threadIdx.x) + 43) % 81)) && (((((int)threadIdx.x) + 43) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 448) / 81) * 49)) + ((((((int)threadIdx.x) + 43) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((9 <= ((((int)threadIdx.x) + 74) % 81)) && (((((int)threadIdx.x) + 74) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 560) / 81) * 49)) + ((((((int)threadIdx.x) + 74) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((9 <= ((((int)threadIdx.x) + 24) % 81)) && (((((int)threadIdx.x) + 24) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 672) / 81) * 49)) + ((((((int)threadIdx.x) + 24) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((9 <= ((((int)threadIdx.x) + 55) % 81)) && (((((int)threadIdx.x) + 55) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 784) / 81) * 49)) + ((((((int)threadIdx.x) + 55) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((9 <= ((((int)threadIdx.x) + 5) % 81)) && (((((int)threadIdx.x) + 5) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 896) / 81) * 49)) + ((((((int)threadIdx.x) + 5) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1008)] = (((((1 <= (((((int)threadIdx.x) / 9) + 4) % 9)) && (((((int)threadIdx.x) + 36) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1008) / 81) * 49)) + ((((((int)threadIdx.x) / 9) + 4) % 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((9 <= ((((int)threadIdx.x) + 67) % 81)) && (((((int)threadIdx.x) + 67) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1120) / 81) * 49)) + ((((((int)threadIdx.x) + 67) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- if (((int)threadIdx.x) < 64) {
- pad_temp_shared[(((int)threadIdx.x) + 1232)] = ((((((int)threadIdx.x) < 55) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1232) / 81) * 49)) + (((((int)threadIdx.x) + 17) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[(((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x))];
- kernel_shared[(((int)threadIdx.x) + 112)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 112) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 224)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 224) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 336) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 448) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 560) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 672) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 784) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 896) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 32256)];
- kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1120) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1232) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1344) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1456) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1568) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1680) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1792) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1904) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 64512)];
- kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2240) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2352)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2352) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2464) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2576)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2576) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2688) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2800)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2800) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2912) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3024)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96768)];
- kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3136) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3248)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3248) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3360)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3360) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3472)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3472) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3584) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3696)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3696) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3808)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3808) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3920)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3920) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 129024)];
- kernel_shared[(((int)threadIdx.x) + 4144)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4144) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4256)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4256) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4368)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4368) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4480) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4592)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4592) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4704)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4704) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4816)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4816) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4928)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4928) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5040)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 161280)];
- kernel_shared[(((int)threadIdx.x) + 5152)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5152) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5264)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5264) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5376)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5376) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5488)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5488) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5600)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5600) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5712)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5712) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5824)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5824) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5936)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5936) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6048)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 193536)];
- kernel_shared[(((int)threadIdx.x) + 6160)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6160) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6272)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6272) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6384)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6384) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6496)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6496) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6608)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6608) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6720)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6720) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6832)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6832) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6944)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6944) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7056)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 225792)];
- kernel_shared[(((int)threadIdx.x) + 7168)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7168) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7280)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7280) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7392)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7392) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7504)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7504) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7616)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7616) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7728)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7728) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7840)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7840) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7952)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7952) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8064)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 258048)];
- kernel_shared[(((int)threadIdx.x) + 8176)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8176) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8288)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8288) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8400)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8400) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8512)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8512) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8624)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8624) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8736)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8736) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8848)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8848) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8960)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8960) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 9072)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 290304)];
- if (((int)threadIdx.x) < 32) {
- kernel_shared[(((int)threadIdx.x) + 9184)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 9184) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 32)] = (((((1 <= ((((((int)threadIdx.x) + 5) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 5) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 32) / 27) * 49)) + ((((((int)threadIdx.x) + 5) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] : 0.0 [...]
+ pad_temp_shared[(((int)threadIdx.x) + 64)] = (((((1 <= ((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 64) / 27) * 49)) + ((((((int)threadIdx.x) + 10) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : [...]
+ pad_temp_shared[(((int)threadIdx.x) + 96)] = (((((1 <= ((((((int)threadIdx.x) + 15) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 15) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 96) / 27) * 49)) + ((((((int)threadIdx.x) + 15) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : [...]
+ pad_temp_shared[(((int)threadIdx.x) + 128)] = (((((1 <= ((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 128) / 27) * 49)) + ((((((int)threadIdx.x) + 20) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] [...]
+ pad_temp_shared[(((int)threadIdx.x) + 160)] = (((((1 <= ((((((int)threadIdx.x) + 25) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 25) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 160) / 27) * 49)) + ((((((int)threadIdx.x) + 25) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] [...]
+ pad_temp_shared[(((int)threadIdx.x) + 192)] = (((((1 <= ((((((int)threadIdx.x) + 3) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 3) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 192) / 27) * 49)) + ((((((int)threadIdx.x) + 3) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0 [...]
+ pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= ((((((int)threadIdx.x) + 8) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 8) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 224) / 27) * 49)) + ((((((int)threadIdx.x) + 8) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0 [...]
+ pad_temp_shared[(((int)threadIdx.x) + 256)] = (((((1 <= ((((((int)threadIdx.x) + 13) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 13) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 256) / 27) * 49)) + ((((((int)threadIdx.x) + 13) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] [...]
+ pad_temp_shared[(((int)threadIdx.x) + 288)] = (((((1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) / 9) + 2) % 3))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) / 9) + 2) % 3)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 288) / 27) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + ((((((int)threadIdx.x) / 9) + 2) % 3) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 320)] = (((((1 <= ((((((int)threadIdx.x) + 23) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 23) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 320) / 27) * 49)) + ((((((int)threadIdx.x) + 23) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] [...]
+ pad_temp_shared[(((int)threadIdx.x) + 352)] = (((((1 <= ((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 352) / 27) * 49)) + ((((((int)threadIdx.x) + 1) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0 [...]
+ pad_temp_shared[(((int)threadIdx.x) + 384)] = (((((1 <= ((((((int)threadIdx.x) + 6) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 6) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 384) / 27) * 49)) + ((((((int)threadIdx.x) + 6) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0 [...]
+ if (((int)threadIdx.x) < 16) {
+ pad_temp_shared[(((int)threadIdx.x) + 416)] = (((((((((int)threadIdx.x) + 11) / 9) + (((int)blockIdx.x) % 7)) < 8) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 416) / 27) * 49)) + (((((int)threadIdx.x) + 11) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
}
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x))];
+ kernel_shared[(((int)threadIdx.x) + 32)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 96)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 160)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 160) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 192) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 224) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 256) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 288)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 9216)];
+ kernel_shared[(((int)threadIdx.x) + 320)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 320) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 352)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 352) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 384) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 416)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 416) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 448) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 480)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 480) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 512) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 544)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 544) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 576)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 18432)];
+ kernel_shared[(((int)threadIdx.x) + 608)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 608) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 640) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 672) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 704)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 704) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 736)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 736) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 768) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 800)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 800) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 832)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 832) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 864)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 27648)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 896) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 928)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 928) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 960) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 992)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 992) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1024) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1056)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1056) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1088) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1120) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 36864)];
+ kernel_shared[(((int)threadIdx.x) + 1184)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1184) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1216) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1248)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1248) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1280) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1312)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1312) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1344) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1376)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1376) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1408) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1440)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 46080)];
+ kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1472) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1504)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1504) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1536) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1568) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1600) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1632)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1632) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1664) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1696)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1696) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 55296)];
+ kernel_shared[(((int)threadIdx.x) + 1760)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1760) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1792) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1824)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1824) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1856) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1888)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1888) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1920) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1952)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1952) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1984) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2048) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2080)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2080) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2112) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2144)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2144) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2176) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2208)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2208) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2240) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2272)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2272) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 73728)];
+ kernel_shared[(((int)threadIdx.x) + 2336)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2336) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2368) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2400)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2400) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2432) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2464) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2496) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2528)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2528) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2560) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2592)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 82944)];
+ kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2624) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2656)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2656) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2688) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2720)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2720) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2752) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2784)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2784) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2816) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2848)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2848) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 92160)];
+ kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2912) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2944) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2976)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2976) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3008) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3040)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3040) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3072)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3072) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3104)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3104) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3136) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3168)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 101376)];
+ kernel_shared[(((int)threadIdx.x) + 3200)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3200) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3232)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3232) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3264)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3264) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3296)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3296) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3328)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3328) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3360)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3360) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3392)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3392) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3424)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3424) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3456)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 110592)];
+ kernel_shared[(((int)threadIdx.x) + 3488)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3488) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3520)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3520) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3552)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3552) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3584) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3616)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3616) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3648)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3648) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3680)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3680) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3712)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3712) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3744)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 119808)];
+ kernel_shared[(((int)threadIdx.x) + 3776)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3776) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3808)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3808) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3840)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3840) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3872)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3872) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3904)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3904) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3936)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3936) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3968)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3968) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4000)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4000) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 129024)];
+ kernel_shared[(((int)threadIdx.x) + 4064)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4064) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4096)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4096) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4128)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 4160)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4160) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4192)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4192) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4224)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4224) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 4256)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4256) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4288)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4288) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4320)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 138240)];
+ kernel_shared[(((int)threadIdx.x) + 4352)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4352) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4384)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4384) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4416)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4416) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 4448)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4448) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4480) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4512)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4512) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 4544)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4544) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4576)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4576) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
__syncthreads();
- for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
- for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
- for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- }
- }
+ for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(ry_outer_inner * 9)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 1)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 27)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 28)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 54)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 55)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 81)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 82)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 108)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 109)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 135)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 136)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 162)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 163)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 189)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 190)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 216)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 217)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 243)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 244)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 270)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 271)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 297)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 298)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 324)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 325)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 351)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 352)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 378)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 379)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 405)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 406)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 1)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 28)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 55)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 82)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 109)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 136)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 163)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 190)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 217)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 244)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 271)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 298)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 325)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 352)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 379)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 406)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 7)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 34)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 61)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 88)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 115)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 142)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 169)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 196)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 223)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 250)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 277)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 304)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 331)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 358)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 385)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 412)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 7)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 8)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 34)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 35)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 61)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 62)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 88)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 89)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 115)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 116)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 142)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 143)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 169)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 170)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 196)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 197)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 223)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 224)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 250)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 251)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 277)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 278)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 304)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 305)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 331)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 332)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 358)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 359)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 385)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 386)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 412)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 413)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
}
}
- for (int i1_inner = 0; i1_inner < 4; ++i1_inner) {
- for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
- compute[(((((((int)blockIdx.x) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[(((((int)blockIdx.x) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner)]), 0.000000e+00f);
- }
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[(((((((int)blockIdx.x) / 7) * 1568) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[(((((int)blockIdx.x) / 7) * 32) + ((int)threadIdx.x))]), 0.000000e+00f);
}
}
@@ -1310,7 +1785,7 @@ 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 38.539 seconds)
+ **Total running time of the script:** ( 5 minutes 42.565 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 31e54a8213..b6e21daf11 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)
- 7.8691 7.8678 7.8797 7.8597 0.0082
+ 7.8563 7.8537 7.8646 7.8505 0.0060
@@ -675,7 +675,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 6.966 seconds)
+ **Total running time of the script:** ( 1 minutes 5.839 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 b992c126ec..eedc323582 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)
- 761.8061 760.6130 765.3606 759.4448 2.5583
+ 758.6698 757.9524 760.1585 757.8985 1.0529
@@ -694,7 +694,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 40.846 seconds)
+ **Total running time of the script:** ( 1 minutes 39.693 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 11db3c1521..09c11c3630 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
@@ -389,25 +389,26 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
@T.prim_func
def main(placeholder: T.Buffer((128, 256), "float32"), placeholder_1: T.Buffer((4916, 16, 1), "float32"), placeholder_2: T.Buffer((4916,), "int32"), placeholder_3: T.Buffer((33,), "int32"), placeholder_4: T.Buffer((128, 512), "float32"), compute: T.Buffer((128, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
- for i0_outer in T.parallel(8):
- compute_1 = T.allocate([256], "float32", "global")
- for i1_outer in range(32):
- compute_2 = T.Buffer((256,), data=compute_1)
- for i_inner_init, j_init in T.grid(16, 16):
- compute_2[i_inner_init * 16 + j_init] = T.float32(0)
- for elem_idx, i_inner, j in T.grid(placeholder_5[i1_outer + 1] - placeholder_5[i1_outer], 16, 16):
+ for i0_outer_i1_outer_fused in T.parallel(64):
+ compute_1 = T.allocate([1024], "float32", "global")
+ compute_2 = T.Buffer((1024,), data=compute_1)
+ for i_outer_inner, nb_j_inner in T.grid(16, 2):
+ for i_inner_init, j_init in T.grid(2, 16):
+ compute_2[i_outer_inner * 64 + i_inner_init * 32 + nb_j_inner * 16 + j_init] = T.float32(0)
+ for elem_idx, i_inner, j in T.grid(T.let(cse_var_1, i0_outer_i1_outer_fused % 16 * 2 + nb_j_inner, placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]), 2, 16):
+ cse_var_1 = T.var("int32")
placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- placeholder_6 = T.Buffer((78656,), data=placeholder_1.data)
- placeholder_7 = T.Buffer((32768,), data=placeholder.data)
- placeholder_8 = T.Buffer((4916,), "int32", data=placeholder_2.data)
- cse_var_1: T.int32 = i_inner * 16 + j
- compute_2[cse_var_1] = compute_2[cse_var_1] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer * 4096 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- for i0_inner in range(16):
- cse_var_2: T.int32 = i0_outer * 8192 + i0_inner * 512 + i1_outer * 16
- compute_3 = T.Buffer((65536,), data=compute.data)
- placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
- compute_3[cse_var_2:cse_var_2 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_2:cse_var_2 + 16], T.Broadcast(T.float32(0), 16))
+ cse_var_3: T.int32 = i0_outer_i1_outer_fused % 16 * 2 + nb_j_inner
+ cse_var_2: T.int32 = i_outer_inner * 64 + i_inner * 32 + nb_j_inner * 16 + j
+ placeholder_6 = T.Buffer((78656,), data=placeholder_1.data)
+ placeholder_7 = T.Buffer((32768,), data=placeholder.data)
+ placeholder_8 = T.Buffer((4916,), "int32", data=placeholder_2.data)
+ compute_2[cse_var_2] = compute_2[cse_var_2] + placeholder_6[placeholder_5[cse_var_3] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer_i1_outer_fused // 16 * 8192 + i_outer_inner * 512 + i_inner * 256 + placeholder_8[placeholder_5[cse_var_3] + elem_idx]], T.float32(0))
+ for i0_inner in range(32):
+ cse_var_4: T.int32 = i0_outer_i1_outer_fused // 16 * 16384 + i0_inner * 512 + i0_outer_i1_outer_fused % 16 * 32
+ compute_3 = T.Buffer((65536,), data=compute.data)
+ placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
+ compute_3[cse_var_4:cse_var_4 + 32] = T.max(compute_2[i0_inner * 32:i0_inner * 32 + 32] + placeholder_5[cse_var_4:cse_var_4 + 32], T.Broadcast(T.float32(0), 32))
@@ -457,7 +458,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 1.522 ms
+ Execution time of this operator: 2.117 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 b983013163..9e695bce91 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
@@ -5,16 +5,16 @@
Computation times
=================
-**00:29.987** total execution time for **how_to_tune_with_autotvm** files:
+**00:43.161** 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``) | 00:29.951 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:43.129 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.023 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.019 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``) | 00:00.005 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``) | 00:00.004 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_arm.py` (``tune_relay_arm.py``) | 00:00.005 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_mobile_gpu.py` (``tune_relay_mobile_gpu.py``) | 00:00.004 | 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 be0f01956a..3389b10200 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
@@ -390,9 +390,9 @@ 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, 8, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2011889
- No: 2 GFLOPS: 124.09/124.09 result: MeasureResult(costs=(0.001865624372881356,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.5087034702301025, timestamp=1674771327.2719243) [('tile_f', [-1, 1, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1803160
- No: 3 GFLOPS: 0.00/124.09 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8495099
+ No: 2 GFLOPS: 90.01/90.01 result: MeasureResult(costs=(0.002572052,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9165120124816895, timestamp=1674803118.8732517) [('tile_f', [-1, 8, 4, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,41072
+ No: 3 GFLOPS: 0.00/90.01 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
@@ -514,10 +514,8 @@ 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, 32, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9198903
- No: 4 GFLOPS: 13.25/124.09 result: MeasureResult(costs=(0.017472539166666665,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.289174795150757, timestamp=1674771329.0655708) [('tile_f', [-1, 1, 16, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2979618
- No: 5 GFLOPS: 213.18/213.18 result: MeasureResult(costs=(0.001085961722972973,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.230830192565918, timestamp=1674771331.4487233) [('tile_f', [-1, 8, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9913017
- No: 6 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7487659
+ No: 4 GFLOPS: 0.00/90.01 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
@@ -639,8 +637,8 @@ 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, 64, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4622248
- No: 7 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9392665
+ No: 5 GFLOPS: 0.00/90.01 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
@@ -762,8 +760,8 @@ 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, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8189408
- No: 8 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2707796
+ No: 6 GFLOPS: 0.00/90.01 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
@@ -885,8 +883,27 @@ 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, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 128, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9900518
- No: 9 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 4, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2806473
+ No: 7 GFLOPS: 164.23/164.23 result: MeasureResult(costs=(0.0014096323918918918,), error_no=MeasureErrorNo.NO_ERROR, all_cost=10.716526985168457, timestamp=1674803132.9284441) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8844285
+ No: 8 GFLOPS: 0.00/164.23 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, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9704202
+ No: 9 GFLOPS: 0.00/164.23 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
@@ -1008,8 +1025,8 @@ 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, 1, 32, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8384330
- No: 10 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 32, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7838471
+ No: 10 GFLOPS: 0.00/164.23 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
@@ -1131,9 +1148,8 @@ 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, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,556949
- No: 11 GFLOPS: 4.01/213.18 result: MeasureResult(costs=(0.0577535055,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6681346893310547, timestamp=1674771334.3050325) [('tile_f', [-1, 8, 2, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,288786
- No: 12 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 512, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,339294
+ No: 11 GFLOPS: 0.00/164.23 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
@@ -1255,8 +1271,9 @@ 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, 4, 32, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3924672
- No: 13 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 256]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10117577
+ No: 12 GFLOPS: 316.20/316.20 result: MeasureResult(costs=(0.0007321365147928994,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7440016269683838, timestamp=1674803135.8509939) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4070442
+ No: 13 GFLOPS: 0.00/316.20 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
@@ -1378,8 +1395,9 @@ 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, 4, 128, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7269291
- No: 14 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7471640
+ No: 14 GFLOPS: 7.72/316.20 result: MeasureResult(costs=(0.02997332,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.854086637496948, timestamp=1674803141.9115317) [('tile_f', [-1, 4, 2, 8]), ('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, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2979825
+ No: 15 GFLOPS: 0.00/316.20 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
@@ -1501,9 +1519,11 @@ 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, 8, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1991594
- No: 15 GFLOPS: 128.52/213.18 result: MeasureResult(costs=(0.001801294857142857,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2695696353912354, timestamp=1674771335.7931848) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6320225
- No: 16 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10253255
+ No: 16 GFLOPS: 111.35/316.20 result: MeasureResult(costs=(0.0020790924415584416,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.28901743888855, timestamp=1674803142.9225328) [('tile_f', [-1, 2, 4, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7074250
+ No: 17 GFLOPS: 58.21/316.20 result: MeasureResult(costs=(0.003976914307692308,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2880961894989014, timestamp=1674803144.4809651) [('tile_f', [-1, 4, 16, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1668736
+ No: 18 GFLOPS: 18.61/316.20 result: MeasureResult(costs=(0.012438851222222222,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.455913782119751, timestamp=1674803145.2864993) [('tile_f', [-1, 8, 1, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6210218
+ No: 19 GFLOPS: 0.00/316.20 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
@@ -1625,8 +1645,8 @@ 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, 1, 8, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7339177
- No: 17 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 2, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9428211
+ No: 20 GFLOPS: 0.00/316.20 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
@@ -1748,376 +1768,7 @@ 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, 8, 2, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2434446
- No: 18 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 128, 1, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9420727
- No: 19 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 128, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1482191
- No: 20 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6155005
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8705647
@@ -2172,9 +1823,9 @@ and measure running time.
Finish loading 20 records
Best config:
- [('tile_f', [-1, 8, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9913017
+ [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4070442
Finish loading 20 records
- Time cost of this operator: 0.001451
+ Time cost of this operator: 0.001051
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 a9ef297bc5..0d8296534b 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 309.6 98.675 (1, 2, 10, 10, 3) 2 1 [309.6]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.115 0.993 (1, 6, 10, 10) 1 1 [3.115]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 1.041 0.332 (1, 1, 10, 10, 3) 1 1 [1.041]
- Total_time - 313.756 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 311.0 98.701 (1, 2, 10, 10, 3) 2 1 [311.0]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.116 0.989 (1, 6, 10, 10) 1 1 [3.116]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.977 0.31 (1, 1, 10, 10, 3) 1 1 [0.977]
+ Total_time - 315.092 - - - - -
@@ -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 136.1 98.006 (1, 6, 10, 10, 1) 2 1 [136.1]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.788 1.288 (1, 6, 10, 10) 1 1 [1.788]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.981 0.706 (1, 1, 10, 10, 3) 1 1 [0.981]
- Total_time - 138.869 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 103.2 97.552 (1, 6, 10, 10, 1) 2 1 [103.2]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.75 1.654 (1, 6, 10, 10) 1 1 [1.75]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.84 0.794 (1, 3, 10, 10, 1) 1 1 [0.84]
+ Total_time - 105.79 - - - - -
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 63b0be738b..e6a9722cc3 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]
61%|###### | 2.09M/3.42M [00:00<00:00, 17.7MB/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 27.6MB/s]
+
0%| | 0.00/3.42M [00:00<?, ?B/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 60.7MB/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 11.795 seconds)
+ **Total running time of the script:** ( 1 minutes 9.814 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 07ecd7f8e0..c0cdd41c5c 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/tmpl3ibk4mc/images/random'
+ '/tmp/tmp6krshqz2/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], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0]
+ :alt: [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [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/tmpl3ibk4mc/images/target contains 8144 images
- /tmp/tmpl3ibk4mc/images/random contains 5000 images
+ /tmp/tmp6krshqz2/images/target contains 8144 images
+ /tmp/tmp6krshqz2/images/random contains 5000 images
@@ -494,13 +494,13 @@ the time on our validation set).
.. code-block:: none
Epoch 1/3
- 328/328 - 48s - loss: 0.2199 - accuracy: 0.9236 - val_loss: 0.1252 - val_accuracy: 0.9588 - 48s/epoch - 146ms/step
+ 328/328 - 47s - loss: 0.2294 - accuracy: 0.9215 - val_loss: 0.1200 - val_accuracy: 0.9596 - 47s/epoch - 144ms/step
Epoch 2/3
- 328/328 - 43s - loss: 0.0936 - accuracy: 0.9659 - val_loss: 0.1593 - val_accuracy: 0.9452 - 43s/epoch - 133ms/step
+ 328/328 - 44s - loss: 0.0962 - accuracy: 0.9643 - val_loss: 0.1094 - val_accuracy: 0.9626 - 44s/epoch - 133ms/step
Epoch 3/3
- 328/328 - 43s - loss: 0.0722 - accuracy: 0.9730 - val_loss: 0.1136 - val_accuracy: 0.9611 - 43s/epoch - 132ms/step
+ 328/328 - 43s - loss: 0.0663 - accuracy: 0.9743 - val_loss: 0.2085 - val_accuracy: 0.9415 - 43s/epoch - 133ms/step
- <keras.callbacks.History object at 0x7f286aec7b90>
+ <keras.callbacks.History object at 0x7fb3666e8e10>
@@ -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 38.806 seconds)
+ **Total running time of the script:** ( 4 minutes 22.236 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 ac85c1f853..5c6308f9c0 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
@@ -5,18 +5,18 @@
Computation times
=================
-**06:56.419** total execution time for **how_to_work_with_microtvm** files:
+**06:36.598** total execution time for **how_to_work_with_microtvm** files:
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:38.806 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:22.236 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:11.795 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:09.814 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:52.416 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 00:51.727 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:09.523 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:09.034 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.878 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:03.787 | 0.0 MB |
+---------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tvmc.py` (``micro_tvmc.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 60f7c42863..3977266a7e 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
=================
-**00:46.429** total execution time for **how_to_work_with_relay** files:
+**00:45.047** 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``) | 00:33.678 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.702 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:11.373 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:10.617 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.372 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.722 | 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 f672959d51..23f36b0dc8 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 0x7f285fef8f80>
+ <function my_cuda_math_rule at 0x7fb3075403b0>
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 c2a2b1e2c9..d9df7ec684 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
@@ -5,18 +5,18 @@
Computation times
=================
-**00:07.531** total execution time for **how_to_work_with_schedules** files:
+**00:06.262** total execution time for **how_to_work_with_schedules** files:
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:04.751 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:03.654 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.378 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``) | 00:01.246 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.590 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.576 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.586 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.559 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.118 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.119 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.051 | 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 f2f7cf3e0a..86da29d8b2 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -340,7 +340,7 @@ The importing needs to happen before the tensorized GEMV being executed.
def main(A: T.Buffer((1024, 64), "float32"), B: T.Buffer((512, 64), "float32"), C: T.Buffer((1024, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
i = T.var("int32")
- T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpavgpahh8/input0.cc'\nsource_filename = \"/tmp/tmpavgpahh8/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 float*, [...]
+ T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpro9643gx/input0.cc'\nsource_filename = \"/tmp/tmpro9643gx/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 float*, [...]
for i, j_outer in T.grid(1024, 32):
T.call_extern("int32", "gemv_update", T.tvm_access_ptr(T.type_annotation("float32"), C.data, i * 512 + j_outer * 16, 16, 2), T.tvm_access_ptr(T.type_annotation("float32"), A.data, i * 64, 64, 1), T.tvm_access_ptr(T.type_annotation("float32"), B.data, j_outer * 1024, 1024, 1), 16, 64, 64)
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 b58aedc59a..f731a317fb 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
=================
-**00:31.068** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:30.483** total execution time for **topic_vta_tutorials_autotvm** files:
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:31.061 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:30.476 | 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 a42222fd27..a885c0cd42 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 33.27s!
+ resnet18_v1 inference graph built in 32.42s!
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 520620c5e1..e8eda5f256 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 22.63s!
+ yolov3-tiny inference graph built in 22.21s!
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 dd03a50c82..6158a274ec 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
=================
-**01:39.889** total execution time for **topic_vta_tutorials_frontend** files:
+**01:37.892** total execution time for **topic_vta_tutorials_frontend** files:
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:50.276 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:49.279 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:49.613 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:48.613 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
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 4bb499960c..67fc38cfa4 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
=================
-**00:03.239** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.145** total execution time for **topic_vta_tutorials_optimize** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.765 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.691 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.474 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.453 | 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 209146a49a..5c9b49b689 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
=================
-**00:00.876** total execution time for **topic_vta_tutorials** files:
+**00:00.816** total execution time for **topic_vta_tutorials** files:
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.480 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.432 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.396 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.384 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index 201a7cf7bc..8b870d5ef8 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -207,13 +207,6 @@ trials, we can load the best schedule from the log file and apply it.
-.. rst-class:: sphx-glr-script-out
-
- .. code-block:: none
-
- *E
-
-
@@ -325,7 +318,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 95.810 ms
+ Execution time of this operator: 95.253 ms
@@ -443,7 +436,7 @@ operations.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 34.940 seconds)
+ **Total running time of the script:** ( 1 minutes 26.964 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 0c519aa892..82a4316e8d 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: 0.78/0.78 result: MeasureResult(costs=(0.3423284788,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.717501878738403, timestamp=1674769767.179735) [('tile_y', [-1, 512]), ('tile_x', [-1, 2])],None,19
- No: 2 GFLOPS: 3.63/3.63 result: MeasureResult(costs=(0.0738693562,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4405498504638672, timestamp=1674769768.6164715) [('tile_y', [-1, 256]), ('tile_x', [-1, 16])],None,48
- No: 3 GFLOPS: 3.68/3.68 result: MeasureResult(costs=(0.0728543994,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4139862060546875, timestamp=1674769770.81386) [('tile_y', [-1, 128]), ('tile_x', [-1, 16])],None,47
- No: 4 GFLOPS: 1.55/3.68 result: MeasureResult(costs=(0.1735699224,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.0015954971313477, timestamp=1674769773.841977) [('tile_y', [-1, 32]), ('tile_x', [-1, 4])],None,25
- No: 5 GFLOPS: 2.56/3.68 result: MeasureResult(costs=(0.10499245719999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9371483325958252, timestamp=1674769776.1143422) [('tile_y', [-1, 2]), ('tile_x', [-1, 16])],None,41
- No: 6 GFLOPS: 2.55/3.68 result: MeasureResult(costs=(0.10533583839999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9270737171173096, timestamp=1674769778.8331282) [('tile_y', [-1, 512]), ('tile_x', [-1, 8])],None,39
- No: 7 GFLOPS: 2.06/3.68 result: MeasureResult(costs=(0.13030431120000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.327479600906372, timestamp=1674769781.949665) [('tile_y', [-1, 256]), ('tile_x', [-1, 4])],None,28
- No: 8 GFLOPS: 9.73/9.73 result: MeasureResult(costs=(0.0275817818,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.9235107898712158, timestamp=1674769782.6539247) [('tile_y', [-1, 2]), ('tile_x', [-1, 128])],None,71
- No: 9 GFLOPS: 13.60/13.60 result: MeasureResult(costs=(0.019740675,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5569474697113037, timestamp=1674769783.323702) [('tile_y', [-1, 16]), ('tile_x', [-1, 512])],None,94
- No: 10 GFLOPS: 1.66/13.60 result: MeasureResult(costs=(0.1617876952,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.7907471656799316, timestamp=1674769786.1661088) [('tile_y', [-1, 2]), ('tile_x', [-1, 2])],None,11
+ No: 1 GFLOPS: 4.15/4.15 result: MeasureResult(costs=(0.0646062716,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2819013595581055, timestamp=1674801589.425423) [('tile_y', [-1, 16]), ('tile_x', [-1, 16])],None,44
+ No: 2 GFLOPS: 3.32/4.15 result: MeasureResult(costs=(0.080842935,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.536332130432129, timestamp=1674801591.742992) [('tile_y', [-1, 8]), ('tile_x', [-1, 8])],None,33
+ No: 3 GFLOPS: 10.64/10.64 result: MeasureResult(costs=(0.0252188724,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6333749294281006, timestamp=1674801592.4050639) [('tile_y', [-1, 1]), ('tile_x', [-1, 512])],None,90
+ No: 4 GFLOPS: 0.86/10.64 result: MeasureResult(costs=(0.310955842,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.192100524902344, timestamp=1674801598.3878272) [('tile_y', [-1, 512]), ('tile_x', [-1, 2])],None,19
+ No: 5 GFLOPS: 14.64/14.64 result: MeasureResult(costs=(0.0183319464,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.8400697708129883, timestamp=1674801599.3476355) [('tile_y', [-1, 64]), ('tile_x', [-1, 64])],None,66
+ No: 6 GFLOPS: 0.51/14.64 result: MeasureResult(costs=(0.52726684,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.658513307571411, timestamp=1674801608.0261073) [('tile_y', [-1, 256]), ('tile_x', [-1, 1])],None,8
+ No: 7 GFLOPS: 1.67/14.64 result: MeasureResult(costs=(0.1606050886,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.800123453140259, timestamp=1674801611.6096852) [('tile_y', [-1, 8]), ('tile_x', [-1, 1])],None,3
+ No: 8 GFLOPS: 13.44/14.64 result: MeasureResult(costs=(0.0199796056,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5835709571838379, timestamp=1674801612.1841033) [('tile_y', [-1, 16]), ('tile_x', [-1, 512])],None,94
+ No: 9 GFLOPS: 0.90/14.64 result: MeasureResult(costs=(0.29902428719999996,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.989312171936035, timestamp=1674801617.2901492) [('tile_y', [-1, 64]), ('tile_x', [-1, 2])],None,16
+ No: 10 GFLOPS: 9.96/14.64 result: MeasureResult(costs=(0.0269493716,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6376514434814453, timestamp=1674801617.9795842) [('tile_y', [-1, 1]), ('tile_x', [-1, 256])],None,80
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index dd429cdd7f..4ef718f65a 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': 513.7583284299944, 'median': 513.1661257500127, 'std': 1.97182240186592}
+ {'mean': 514.6029234700006, 'median': 515.1770631500028, 'std': 1.6195756103389116}
@@ -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: 14.57/ 15.10 GFLOPS | Progress: (4/20) | 7.70 s
[Task 1/25] Current/Best: 23.19/ 23.19 GFLOPS | Progress: (8/20) | 10.59 s
[Task 1/25] Current/Best: 5.65/ 23.19 GFLOPS | Progress: (12/20) | 16.91 s
[Task 1/25] Current/Best: 10.49/ 23.19 GFLOPS | Progress: (16/20) | 19.75 s
[Task 1/25] Current/Best: 19.15/ 23.19 GFLOPS | Progress: (20/20) | 23.65 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 18.91/ 19.64 GFLOPS | Progress: (4/20) | 3.83 s
[Task 2/25] Current/Best: 13.57/ 19.64 GFLOPS | Progress: (8/20) | 5.91 s
[Task 2/25] Current/Best: 17.41/ 19.64 GFLOPS | Progress: (12/20) | 7.47 s
[Task 2/25] Current/Best: 12.97/ 19.64 GFLOPS | Progress: (16/20) | 9.47 s
[Task 2/25] Current/Best: 7.99/ 19.64 GFLOPS | Progress: (20/20) | 11.04 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 13.19/ 16.34 GFLOPS | Progress: (4/20) | 4.12 s
[Task 3/25] Current/Best: 9.79/ 19.89 GFLOPS | Progress: (8/20) | 6.22 s
[Task 3/25] Current/Best: 7.30/ 19.89 GFLOPS | Progress: (12/20) | 9.34 s
[Task 3/25] Current/Best: 12.17/ 19.89 GFLOPS | Progress: (16/20) | 11.78 s
[Task 3/25] Current/Best: 16.77/ 19.89 GFLOPS | Progress: (20/20) | 13.86 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 9.43/ 16.41 GFLOPS | Progress: (4/20) | 7.99 s
[Task 4/25] Current/Best: 10.61/ 20.63 GFLOPS | Progress: (8/20) | 10.02 s
[Task 4/25] Current/Best: 21.86/ 21.86 GFLOPS | Progress: (12/20) | 16.68 s
[Task 4/25] Current/Best: 11.43/ 21.86 GFLOPS | Progress: (16/20) | 19.31 s
[Task 4/25] Current/Best: 16.97/ 21.86 GFLOPS | Progress: (20/20) | 21.62 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 5.78/ 17.71 GFLOPS | Progress: (4/20) | 5.43 s
[Task 5/25] Current/Best: 20.51/ 20.51 GFLOPS | Progress: (8/20) | 7.47 s
[Task 5/25] Current/Best: 6.91/ 20.51 GFLOPS | Progress: (12/20) | 9.47 s
[Task 5/25] Current/Best: 10.11/ 20.51 GFLOPS | Progress: (16/20) | 11.42 s
[Task 5/25] Current/Best: 13.73/ 20.51 GFLOPS | Progress: (20/20) | 13.47 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 11.50/ 14.50 GFLOPS | Progress: (4/20) | 4.72 s
[Task 6/25] Current/Best: 15.07/ 15.07 GFLOPS | Progress: (8/20) | 7.66 s
[Task 6/25] Current/Best: 14.59/ 15.07 GFLOPS | Progress: (12/20) | 11.63 s
[Task 6/25] Current/Best: 13.68/ 15.07 GFLOPS | Progress: (16/20) | 14.59 s
[Task 6/25] Current/Best: 8.70/ 15.07 GFLOPS | Progress: (20/20) | 18.66 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 3.15/ 18.96 GFLOPS | Progress: (4/20) | 4.50 s
[Task 7/25] Current/Best: 15.14/ 22.79 GFLOPS | Progress: (8/20) | 7.50 s
[Task 7/25] Current/Best: 16.16/ 22.79 GFLOPS | Progress: (12/20) | 9.85 s
[Task 7/25] Current/Best: 14.86/ 22.79 GFLOPS | Progress: (16/20) | 12.21 s
[Task 7/25] Current/Best: 10.94/ 22.79 GFLOPS | Progress: (20/20) | 14.44 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 19.58/ 19.58 GFLOPS | Progress: (4/20) | 3.74 s
[Task 8/25] Current/Best: 16.13/ 19.58 GFLOPS | Progress: (8/20) | 7.79 s
[Task 8/25] Current/Best: 21.04/ 21.04 GFLOPS | Progress: (12/20) | 11.44 s
[Task 8/25] Current/Best: 2.95/ 21.04 GFLOPS | Progress: (16/20) | 14.78 s
[Task 8/25] Current/Best: 17.07/ 21.04 GFLOPS | Progress: (20/20) | 20.16 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 9.05/ 17.22 GFLOPS | Progress: (4/20) | 9.89 s
[Task 9/25] Current/Best: 15.20/ 18.28 GFLOPS | Progress: (8/20) | 11.51 s
[Task 9/25] Current/Best: 6.61/ 18.28 GFLOPS | Progress: (12/20) | 22.76 s
[Task 9/25] Current/Best: 12.40/ 18.48 GFLOPS | Progress: (16/20) | 25.40 s
[Task 9/25] Current/Best: 12.91/ 20.48 GFLOPS | Progress: (20/20) | 27.05 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 12.27/ 16.36 GFLOPS | Progress: (4/20) | 5.25 s
[Task 10/25] Current/Best: 14.15/ 16.36 GFLOPS | Progress: (8/20) | 7.23 s
[Task 10/25] Current/Best: 15.64/ 18.90 GFLOPS | Progress: (12/20) | 8.93 s
[Task 10/25] Current/Best: 2.61/ 18.90 GFLOPS | Progress: (16/20) | 11.57 s
[Task 10/25] Current/Best: 6.30/ 18.90 GFLOPS | Progress: (20/20
) | 13.29 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 11.03/ 16.71 GFLOPS | Progress: (4/20) | 5.15 s
[Task 11/25] Current/Best: 17.89/ 17.89 GFLOPS | Progress: (8/20) | 8.37 s
[Task 11/25] Current/Best: 18.61/ 18.61 GFLOPS | Progress: (12/20) | 10.77 s
[Task 11/25] Current/Best: 19.92/ 19.92 GFLOPS | Progress: (16/20) | 13.58 s
[Task 11/25] Current/Best: 9.78/ 19.92 GFLOPS | Progress: (20/20) | 16.15 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 15.86/ 18.21 GFLOPS | Progress: (4/20) | 4.72 s
[Task 12/25] Current/Best: 8.35/ 18.21 GFLOPS | Progress: (8/20) | 6.77 s
[Task 12/25] Current/Best: 10.83/ 18.21 GFLOPS | Progress: (12/20) | 10.25 s
[Task 12/25] Current/Best: 11.89/ 18.21 GFLOPS | Progress: (16/20) | 12.54 s
[Task 12/25] Current/Best: 8.87/ 18.21 GFLOPS | Progress: (20/20) | 15.58 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 13.30/ 13.30 GFLOPS | Progress: (4/20) | 6.93 s
[Task 13/25] Current/Best: 20.01/ 20.01 GFLOPS | Progress: (8/20) | 9.91 s
[Task 13/25] Current/Best: 6.78/ 20.01 GFLOPS | Progress: (12/20) | 13.01 s
[Task 13/25] Current/Best: 8.06/ 20.01 GFLOPS | Progress: (16/20) | 15.32 s
[Task 13/25] Current/Best: 8.21/ 20.01 GFLOPS | Progress: (20/20) | 17.99 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 9.25/ 15.66 GFLOPS | Progress: (4/20) | 5.42 s
[Task 14/25] Current/Best: 10.79/ 15.66 GFLOPS | Progress: (8/20) | 11.51 s
[Task 14/25] Current/Best: 17.35/ 17.35 GFLOPS | Progress: (12/20) | 15.08 s
[Task 14/25] Current/Best: 12.52/ 17.35 GFLOPS | Progress: (16/20) | 19.85 s
[Task 14/25] Current/Best: 13.23/ 20.21 GFLOPS | Progress: (20/20) | 23.31 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 12.08/ 13.55 GFLOPS | Progress: (4/20) | 5.20 s
[Task 15/25] Current/Best: 17.70/ 17.70 GFLOPS | Progress: (8/20) | 8.03 s
[Task 15/25] Current/Best: 17.06/ 17.70 GFLOPS | Progress: (12/20) | 10.26 s Done.
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 9.47/ 9.47 GFLOPS | Progress: (4/20) | 8.71 s
[Task 1/25] Current/Best: 12.87/ 22.12 GFLOPS | Progress: (8/20) | 11.93 s
[Task 1/25] Current/Best: 6.39/ 22.12 GFLOPS | Progress: (12/20) | 14.38 s
[Task 1/25] Current/Best: 13.19/ 23.54 GFLOPS | Progress: (16/20) | 17.33 s
[Task 1/25] Current/Best: 17.76/ 23.54 GFLOPS | Progress: (20/20) | 20.02 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 21.62/ 21.62 GFLOPS | Progress: (4/20) | 3.53 s
[Task 2/25] Current/Best: 6.22/ 21.62 GFLOPS | Progress: (8/20) | 4.80 s
[Task 2/25] Current/Best: 16.06/ 21.62 GFLOPS | Progress: (12/20) | 7.87 s
[Task 2/25] Current/Best: 13.57/ 21.62 GFLOPS | Progress: (16/20) | 9.23 s
[Task 2/25] Current/Best: 14.14/ 21.62 GFLOPS | Progress: (20/20) | 10.82 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 18.26/ 18.33 GFLOPS | Progress: (4/20) | 4.59 s
[Task 3/25] Current/Best: 20.19/ 21.03 GFLOPS | Progress: (8/20) | 7.63 s
[Task 3/25] Current/Best: 17.57/ 21.03 GFLOPS | Progress: (12/20) | 9.81 s
[Task 3/25] Current/Best: 20.10/ 21.03 GFLOPS | Progress: (16/20) | 11.95 s
[Task 3/25] Current/Best: 9.06/ 21.03 GFLOPS | Progress: (20/20) | 14.14 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 13.80/ 16.00 GFLOPS | Progress: (4/20) | 6.65 s
[Task 4/25] Current/Best: 6.58/ 18.65 GFLOPS | Progress: (8/20) | 8.48 s
[Task 4/25] Current/Best: 11.70/ 18.65 GFLOPS | Progress: (12/20) | 10.53 s
[Task 4/25] Current/Best: 19.67/ 19.67 GFLOPS | Progress: (16/20) | 14.90 s
[Task 4/25] Current/Best: 10.30/ 19.67 GFLOPS | Progress: (20/20) | 18.28 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 4.89/ 10.76 GFLOPS | Progress: (4/20) | 4.77 s
[Task 5/25] Current/Best: 15.42/ 15.42 GFLOPS | Progress: (8/20) | 6.90 s
[Task 5/25] Current/Best: 17.54/ 18.47 GFLOPS | Progress: (12/20) | 8.79 s
[Task 5/25] Current/Best: 9.04/ 18.47 GFLOPS | Progress: (16/20) | 11.18 s
[Task 5/25] Current/Best: 4.44/ 18.47 GFLOPS | Progress: (20/20) | 15.08 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 11.03/ 11.03 GFLOPS | Progress: (4/20) | 5.14 s
[Task 6/25] Current/Best: 14.02/ 18.26 GFLOPS | Progress: (8/20) | 7.53 s
[Task 6/25] Current/Best: 17.79/ 18.26 GFLOPS | Progress: (12/20) | 9.90 s
[Task 6/25] Current/Best: 5.29/ 21.69 GFLOPS | Progress: (16/20) | 13.02 s
[Task 6/25] Current/Best: 12.38/ 21.69 GFLOPS | Progress: (20/20) | 16.20 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 4.63/ 18.55 GFLOPS | Progress: (4/20) | 4.41 s
[Task 7/25] Current/Best: 7.29/ 18.55 GFLOPS | Progress: (8/20) | 7.12 s
[Task 7/25] Current/Best: 5.20/ 18.55 GFLOPS | Progress: (12/20) | 9.94 s
[Task 7/25] Current/Best: 6.03/ 18.67 GFLOPS | Progress: (16/20) | 12.45 s
[Task 7/25] Current/Best: 5.05/ 18.67 GFLOPS | Progress: (20/20) | 15.55 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 7.28/ 15.78 GFLOPS | Progress: (4/20) | 13.86 s
[Task 8/25] Current/Best: 13.39/ 15.78 GFLOPS | Progress: (8/20) | 18.04 s
[Task 8/25] Current/Best: 4.88/ 21.13 GFLOPS | Progress: (12/20) | 23.41 s
[Task 8/25] Current/Best: 17.34/ 21.13 GFLOPS | Progress: (16/20) | 28.29 s
[Task 8/25] Current/Best: 3.07/ 21.13 GFLOPS | Progress: (20/20) | 35.70 s
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 20.83/ 20.83 GFLOPS | Progress: (4/20) | 3.65 s
[Task 9/25] Current/Best: 15.94/ 20.83 GFLOPS | Progress: (8/20) | 14.37 s
[Task 9/25] Current/Best: 9.94/ 20.83 GFLOPS | Progress: (12/20) | 17.27 s
[Task 9/25] Current/Best: 8.02/ 20.83 GFLOPS | Progress: (16/20) | 19.03 s
[Task 9/25] Current/Best: 10.14/ 21.01 GFLOPS | Progress: (20
/20) | 22.43 s
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 13.82/ 14.09 GFLOPS | Progress: (4/20) | 3.83 s
[Task 10/25] Current/Best: 11.05/ 18.14 GFLOPS | Progress: (8/20) | 6.17 s
[Task 10/25] Current/Best: 15.25/ 18.14 GFLOPS | Progress: (12/20) | 8.63 s
[Task 10/25] Current/Best: 14.14/ 18.14 GFLOPS | Progress: (16/20) | 10.97 s
[Task 10/25] Current/Best: 10.04/ 18.14 GFLOPS | Progress: (20/20) | 13.92 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 12.36/ 19.74 GFLOPS | Progress: (4/20) | 5.42 s
[Task 11/25] Current/Best: 7.01/ 19.74 GFLOPS | Progress: (8/20) | 8.20 s
[Task 11/25] Current/Best: 11.38/ 19.74 GFLOPS | Progress: (12/20) | 10.89 s
[Task 11/25] Current/Best: 15.68/ 19.74 GFLOPS | Progress: (16/20) | 13.15 s
[Task 11/25] Current/Best: 9.64/ 19.74 GFLOPS | Progress: (20/20) | 16.46 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 6.76/ 18.48 GFLOPS | Progress: (4/20) | 4.09 s Done.
Done.
-
[Task 15/25] Current/Best: 11.63/ 17.75 GFLOPS | Progress: (16/20) | 13.58 s
[Task 15/25] Current/Best: 19.83/ 19.83 GFLOPS | Progress: (20/20) | 15.06 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 6.48/ 16.07 GFLOPS | Progress: (4/20) | 4.23 s
[Task 16/25] Current/Best: 17.35/ 18.77 GFLOPS | Progress: (8/20) | 5.86 s
[Task 16/25] Current/Best: 19.78/ 19.78 GFLOPS | Progress: (12/20) | 8.75 s
[Task 16/25] Current/Best: 7.01/ 19.78 GFLOPS | Progress: (16/20) | 10.50 s
[Task 16/25] Current/Best: 20.59/ 20.59 GFLOPS | Progress: (20/20) | 12.72 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 11.82/ 14.58 GFLOPS | Progress: (4/20) | 4.50 s
[Task 17/25] Current/Best: 16.45/ 18.36 GFLOPS | Progress: (8/20) | 6.96 s
[Task 17/25] Current/Best: 6.20/ 18.36 GFLOPS | Progress: (12/20) | 9.53 s
[Task 17/25] Current/Best: 19.23/ 19.23 GFLOPS | Progress: (16/20) | 12.65 s
[Task 17/25] Current/Best: 15.32/ 19.23 GFLOPS | Progress: (20/20) | 14.98 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 13.91/ 13.91 GFLOPS | Progress: (4/20) | 4.93 s
[Task 18/25] Current/Best: 7.12/ 17.70 GFLOPS | Progress: (8/20) | 7.14 s
[Task 18/25] Current/Best: 3.06/ 17.70 GFLOPS | Progress: (12/20) | 11.34 s
[Task 18/25] Current/Best: 14.07/ 17.70 GFLOPS | Progress: (16/20) | 14.05 s
[Task 18/25] Current/Best: 9.50/ 19.63 GFLOPS | Progress: (20/20) | 16.82 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 18.16/ 18.16 GFLOPS | Progress: (4/20) | 5.01 s
[Task 19/25] Current/Best: 12.05/ 19.44 GFLOPS | Progress: (8/20) | 7.77 s
[Task 19/25] Current/Best: 16.57/ 19.44 GFLOPS | Progress: (12/20) | 12.84 s
[Task 19/25] Current/Best: 11.57/ 19.44 GFLOPS | Progress: (16/20) | 15.60 s
[Task 19/25] Current/Best: 11.24/ 19.75 GFLOPS | Progress: (20/20) | 18.04 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 1.56/ 17.09 GFLOPS | Progress: (4/20) | 5.38 s
[Task 20/25] Current/Best: 14.64/ 17.09 GFLOPS | Progress: (8/20) | 9.53 s
[Task 20/25] Current/Best: 18.19/ 18.19 GFLOPS | Progress: (12/20) | 11.90 s
[Task 20/25] Current/Best: 12.09/ 18.19 GFLOPS | Progress: (16/20) | 13.32 s
[Task 20/25] Current/Best: 9.10/ 18.19 GFLOPS | Progress: (20/20) | 17.73 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 12.76/ 15.73 GFLOPS | Progress: (4/20) | 4.52 s
[Task 21/25] Current/Best: 22.56/ 22.56 GFLOPS | Progress: (8/20) | 7.88 s
[Task 21/25] Current/Best: 19.76/ 22.56 GFLOPS | Progress: (12/20) | 9.50 s
[Task 21/25] Current/Best: 14.03/ 22.56 GFLOPS | Progress: (16/20) | 11.15 s Done.
+
[Task 12/25] Current/Best: 12.97/ 18.48 GFLOPS | Progress: (8/20) | 6.93 s
[Task 12/25] Current/Best: 12.57/ 18.48 GFLOPS | Progress: (12/20) | 11.16 s
[Task 12/25] Current/Best: 9.03/ 18.48 GFLOPS | Progress: (16/20) | 15.62 s
[Task 12/25] Current/Best: 11.79/ 18.48 GFLOPS | Progress: (20/20) | 18.94 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 15.09/ 15.09 GFLOPS | Progress: (4/20) | 4.68 s
[Task 13/25] Current/Best: 3.11/ 21.02 GFLOPS | Progress: (8/20) | 8.20 s
[Task 13/25] Current/Best: 21.98/ 21.98 GFLOPS | Progress: (12/20) | 12.61 s
[Task 13/25] Current/Best: 13.40/ 21.98 GFLOPS | Progress: (16/20) | 15.40 s
[Task 13/25] Current/Best: 3.10/ 21.98 GFLOPS | Progress: (20/20) | 18.85 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 10.42/ 21.12 GFLOPS | Progress: (4/20) | 6.38 s
[Task 14/25] Current/Best: 1.60/ 21.12 GFLOPS | Progress: (8/20) | 9.59 s
[Task 14/25] Current/Best: 3.93/ 22.33 GFLOPS | Progress: (12/20) | 12.29 s
[Task 14/25] Current/Best: 6.97/ 22.33 GFLOPS | Progress: (16/20) | 15.58 s
[Task 14/25] Current/Best: 11.87/ 22.33 GFLOPS | Progress: (20/20) | 20.47 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 18.53/ 18.53 GFLOPS | Progress: (4/20) | 7.66 s
[Task 15/25] Current/Best: 16.32/ 18.53 GFLOPS | Progress: (8/20) | 9.98 s
[Task 15/25] Current/Best: 11.76/ 18.53 GFLOPS | Progress: (12/20) | 12.19 s Done.
+
[Task 15/25] Current/Best: 18.69/ 18.69 GFLOPS | Progress: (16/20) | 14.25 s
[Task 15/25] Current/Best: 14.78/ 18.69 GFLOPS | Progress: (20/20) | 17.28 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 16.43/ 18.42 GFLOPS | Progress: (4/20) | 4.25 s
[Task 16/25] Current/Best: 13.88/ 18.42 GFLOPS | Progress: (8/20) | 6.43 s
[Task 16/25] Current/Best: 20.85/ 20.85 GFLOPS | Progress: (12/20) | 9.61 s
[Task 16/25] Current/Best: 10.90/ 20.85 GFLOPS | Progress: (16/20) | 12.30 s
[Task 16/25] Current/Best: 9.58/ 20.85 GFLOPS | Progress: (20/20) | 14.32 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 12.81/ 23.57 GFLOPS | Progress: (4/20) | 3.98 s
[Task 17/25] Current/Best: 19.51/ 23.57 GFLOPS | Progress: (8/20) | 7.47 s
[Task 17/25] Current/Best: 11.28/ 23.57 GFLOPS | Progress: (12/20) | 11.33 s
[Task 17/25] Current/Best: 18.24/ 23.57 GFLOPS | Progress: (16/20) | 13.60 s
[Task 17/25] Current/Best: 14.11/ 23.57 GFLOPS | Progress: (20/20) | 16.58 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 18.96/ 18.96 GFLOPS | Progress: (4/20) | 5.87 s
[Task 18/25] Current/Best: 12.04/ 18.96 GFLOPS | Progress: (8/20) | 8.63 s
[Task 18/25] Current/Best: 6.34/ 19.05 GFLOPS | Progress: (12/20) | 10.90 s
[Task 18/25] Current/Best: 21.23/ 21.23 GFLOPS | Progress: (16/20) | 17.47 s
[Task 18/25] Current/Best: 13.28/ 21.23 GFLOPS | Progress: (20/20) | 19.77 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 10.24/ 12.04 GFLOPS | Progress: (4/20) | 6.07 s
[Task 19/25] Current/Best: 18.29/ 18.29 GFLOPS | Progress: (8/20) | 9.51 s
[Task 19/25] Current/Best: 12.01/ 18.43 GFLOPS | Progress: (12/20) | 11.65 s
[Task 19/25] Current/Best: 18.19/ 19.07 GFLOPS | Progress: (16/20) | 15.42 s
[Task 19/25] Current/Best: 18.05/ 21.79 GFLOPS | Progress: (20/20) | 19.15 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 16.20/ 19.51 GFLOPS | Progress: (4/20) | 4.11 s
[Task 20/25] Current/Best: 12.14/ 20.08 GFLOPS | Progress: (8/20) | 8.48 s
[Task 20/25] Current/Best: 4.06/ 20.08 GFLOPS | Progress: (12/20) | 10.97 s
[Task 20/25] Current/Best: 5.47/ 20.08 GFLOPS | Progress: (16/20) | 14.30 s
[Task 20/25] Current/Best: 10.39/ 20.08 GFLOPS | Progress: (20/20) | 17.48 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 15.27/ 19.07 GFLOPS | Progress: (4/20) | 3.60 s
[Task 21/25] Current/Best: 16.82/ 19.07 GFLOPS | Progress: (8/20) | 4.92 s
[Task 21/25] Current/Best: 9.18/ 20.99 GFLOPS | Progress: (12/20) | 8.66 s
[Task 21/25] Current/Best: 14.25/ 20.99 GFLOPS | Progress: (16/20) | 10.98 s
[Task 21/25] Current/Best: 10.85/ 20.99 GFLOPS | Progress: (20/20)
| 13.32 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
Done.
-
[Task 21/25] Current/Best: 18.22/ 22.56 GFLOPS | Progress: (20/20) | 12.87 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 19.37/ 19.37 GFLOPS | Progress: (4/20) | 5.11 s
[Task 22/25] Current/Best: 9.58/ 19.93 GFLOPS | Progress: (8/20) | 6.88 s
[Task 22/25] Current/Best: 19.77/ 19.93 GFLOPS | Progress: (12/20) | 8.80 s
[Task 22/25] Current/Best: 2.69/ 19.93 GFLOPS | Progress: (16/20) | 12.20 s
[Task 22/25] Current/Best: 13.69/ 19.93 GFLOPS | Progress: (20/20) | 14.64 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 16.36/ 18.13 GFLOPS | Progress: (4/20) | 4.43 s
[Task 23/25] Current/Best: 18.91/ 18.91 GFLOPS | Progress: (8/20) | 6.71 s
[Task 23/25] Current/Best: 19.06/ 22.84 GFLOPS | Progress: (12/20) | 9.01 s
[Task 23/25] Current/Best: 14.44/ 22.84 GFLOPS | Progress: (16/20) | 11.84 s
[Task 23/25] Current/Best: 11.65/ 22.84 GFLOPS | Progress: (20/20) | 15.20 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 5.43/ 7.98 GFLOPS | Progress: (4/20) | 5.54 s
[Task 24/25] Current/Best: 9.92/ 9.92 GFLOPS | Progress: (8/20) | 16.47 s
[Task 24/25] Current/Best: 8.90/ 9.92 GFLOPS | Progress: (12/20) | 28.36 s
[Task 24/25] Current/Best: 5.26/ 9.92 GFLOPS | Progress: (16/20) | 32.86 s
[Task 24/25] Current/Best: 1.98/ 9.92 GFLOPS | Progress: (20/20) | 43.22 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 3.04/ 7.66 GFLOPS | Progress: (4/20) | 5.96 s Done.
-
[Task 25/25] Current/Best: 5.26/ 7.66 GFLOPS | Progress: (8/20) | 16.62 s
[Task 25/25] Current/Best: 7.77/ 8.09 GFLOPS | Progress: (12/20) | 27.54 s
[Task 25/25] Current/Best: 9.10/ 9.10 GFLOPS | Progress: (16/20) | 30.29 s
[Task 25/25] Current/Best: 2.83/ 9.10 GFLOPS | Progress: (20/20) | 36.42 s
+
[Task 22/25] Current/Best: 18.30/ 18.30 GFLOPS | Progress: (4/20) | 3.69 s
[Task 22/25] Current/Best: 8.91/ 18.30 GFLOPS | Progress: (8/20) | 6.12 s
[Task 22/25] Current/Best: 7.99/ 18.30 GFLOPS | Progress: (12/20) | 8.09 s
[Task 22/25] Current/Best: 12.03/ 18.30 GFLOPS | Progress: (16/20) | 11.39 s
[Task 22/25] Current/Best: 14.29/ 18.30 GFLOPS | Progress: (20/20) | 14.35 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 14.01/ 17.88 GFLOPS | Progress: (4/20) | 8.93 s
[Task 23/25] Current/Best: 9.69/ 17.88 GFLOPS | Progress: (8/20) | 12.02 s
[Task 23/25] Current/Best: 23.63/ 23.63 GFLOPS | Progress: (12/20) | 14.18 s
[Task 23/25] Current/Best: 22.91/ 23.63 GFLOPS | Progress: (16/20) | 16.35 s
[Task 23/25] Current/Best: 19.92/ 23.63 GFLOPS | Progress: (20/20) | 21.68 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 5.25/ 9.19 GFLOPS | Progress: (4/20) | 4.46 s
[Task 24/25] Current/Best: 2.81/ 10.20 GFLOPS | Progress: (8/20) | 15.39 s
[Task 24/25] Current/Best: 2.18/ 10.20 GFLOPS | Progress: (12/20) | 24.52 s
[Task 24/25] Current/Best: 2.27/ 10.20 GFLOPS | Progress: (16/20) | 35.18 s
[Task 24/25] Current/Best: 5.54/ 10.20 GFLOPS | Progress: (20/20) | 46.11 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 25/25] Current/Best: 1.52/ 8.88 GFLOPS | Progress: (4/20) | 13.68 s Done.
+
[Task 25/25] Current/Best: 1.52/ 8.88 GFLOPS | Progress: (8/20) | 17.26 s
[Task 25/25] Current/Best: 5.77/ 8.88 GFLOPS | Progress: (12/20) | 27.93 s
[Task 25/25] Current/Best: 1.52/ 8.88 GFLOPS | Progress: (16/20) | 31.27 s
[Task 25/25] Current/Best: 1.55/ 8.88 GFLOPS | Progress: (20/20) | 41.98 s
@@ -664,8 +664,8 @@ Verify that the optimized model runs and produces the same results:
.. code-block:: none
- class='n02123045 tabby, tabby cat' with probability=0.621104
- class='n02123159 tiger cat' with probability=0.356378
+ class='n02123045 tabby, tabby cat' with probability=0.621103
+ class='n02123159 tiger cat' with probability=0.356379
class='n02124075 Egyptian cat' with probability=0.019712
class='n02129604 tiger, Panthera tigris' with probability=0.001215
class='n04040759 radiator' with probability=0.000262
@@ -722,8 +722,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 428.178402379981, 'median': 427.21720610002194, 'std': 3.5006333000484164}
- unoptimized: {'mean': 513.7583284299944, 'median': 513.1661257500127, 'std': 1.97182240186592}
+ optimized: {'mean': 423.5321302300008, 'median': 422.38251159999436, 'std': 3.261529146558627}
+ unoptimized: {'mean': 514.6029234700006, 'median': 515.1770631500028, 'std': 1.6195756103389116}
@@ -746,7 +746,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 11 minutes 35.752 seconds)
+ **Total running time of the script:** ( 12 minutes 6.103 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 e859908c26..f02bc2b63e 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
- 2.723e-07 secs/op
+ 1.4e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index 9d1173cb79..9c4f041bc4 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -277,7 +277,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x20e93390)), stage(b, placeholder(b, 0xcec3270)), 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. [...]
+ [stage(a, placeholder(a, 0xd680740)), stage(b, placeholder(b, 0x88632f0)), 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 90c05186cb..c3ba992fec 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,31 +5,31 @@
Computation times
=================
-**15:20.112** total execution time for **tutorial** files:
+**15:46.329** total execution time for **tutorial** files:
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 11:35.752 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 12:06.103 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:34.940 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:26.964 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:02.863 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:01.053 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:35.878 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:35.691 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:29.043 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:34.308 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.840 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.230 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:00.620 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.824 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.177 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.157 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_uma.py` (``uma.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_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_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 ea48fab23d..0b88783826 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -285,7 +285,7 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
- Numpy running time: 0.000007
+ Numpy running time: 0.000008
naive: 0.000007
@@ -504,10 +504,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 7.45827000173449e-06 1.0
- naive 6.742099999999999e-06 0.9039763910976756
- parallel 7.1921999999999995e-06 0.9643255068973617
- vector 2.45441e-05 3.290856994221454
+ numpy 7.67368999959217e-06 1.0
+ naive 6.672500000000001e-06 0.869529522349042
+ parallel 6.9357e-06 0.9038285362542151
+ vector 2.45506e-05 3.199321317554498
@@ -928,7 +928,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.019207
+ Numpy running time: 0.018616
@@ -986,7 +986,7 @@ optimizations.
.. code-block:: none
- none: 3.556046
+ none: 3.410544
@@ -1086,7 +1086,7 @@ schedule.
.. code-block:: none
- blocking: 0.290802
+ blocking: 0.307069
@@ -1170,7 +1170,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.335537
+ vectorization: 0.336873
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1236,7 +1236,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.119023
+ loop permutation: 0.116249
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1327,7 +1327,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.109151
+ array packing: 0.107829
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1410,7 +1410,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.110830
+ block caching: 0.110483
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1484,7 +1484,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.146173
+ parallelization: 0.146349
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1554,13 +1554,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.5560458481000006 1.0
- blocking 0.29080179549999996 0.08177672840055639
- vectorization 0.3355366229 0.09435666389939197
- loop permutation 0.11902299029999999 0.03347060060083143
- array packing 0.1091505121 0.03069434893768854
- block caching 0.11083041869999999 0.03116675752626103
- parallelization 0.1461727953 0.041105430453912815
+ none 3.4105436585 1.0
+ blocking 0.3070694962 0.09003535123636361
+ vectorization 0.3368725663 0.09877386130519754
+ loop permutation 0.1162493442 0.03408528253560839
+ array packing 0.10782897310000002 0.031616359119538306
+ block caching 0.1104832868 0.032394626154292346
+ parallelization 0.14634923719999998 0.042910823567749376
@@ -1602,7 +1602,7 @@ the computation for specific platforms.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 2.863 seconds)
+ **Total running time of the script:** ( 1 minutes 1.053 seconds)
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
diff --git a/docs/commit_hash b/docs/commit_hash
index ca0a15b00d..ccfcbde23c 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-4ddb87631b430a6b29e6075adaa27ae4fa90598e
+18b7dc1dd928bd3e621ff68ad07b354974be13a8
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index ad65b9b19b..c7a9d079a5 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 17.727 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 14.131 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 a589b534f7..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 989ms/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 f833d80ceb..6e75c9dc48 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.zipcd4289a4-40e9-4fab-8b7e-a92cf4b3ba76 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.zip28c6522a-69df-4ac8-ab68-ff22f5a4cbfc 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 b3d16de63f..9db54b4c22 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -449,13 +449,11 @@ 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, 54.4MB/s]
- 28%|##7 | 11.5M/41.5M [00:00<00:01, 31.4MB/s]
- 39%|###8 | 16.0M/41.5M [00:00<00:00, 32.1MB/s]
- 58%|#####7 | 24.0M/41.5M [00:00<00:00, 44.9MB/s]
- 81%|######## | 33.4M/41.5M [00:00<00:00, 60.0MB/s]
- 96%|#########6| 39.9M/41.5M [00:00<00:00, 45.3MB/s]
-100%|##########| 41.5M/41.5M [00:00<00:00, 44.1MB/s]
+ 19%|#9 | 7.99M/41.5M [00:00<00:00, 77.5MB/s]
+ 39%|###8 | 16.0M/41.5M [00:00<00:00, 75.0MB/s]
+ 58%|#####7 | 24.0M/41.5M [00:00<00:00, 66.7MB/s]
+ 90%|########9 | 37.2M/41.5M [00:00<00:00, 85.3MB/s]
+100%|##########| 41.5M/41.5M [00:00<00:00, 78.9MB/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 a93d1ef9dc..bf244bce64 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -432,10 +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]
- 36%|###5 | 16.0M/44.7M [00:00<00:00, 139MB/s]
- 66%|######5 | 29.3M/44.7M [00:00<00:00, 136MB/s]
- 95%|#########4| 42.3M/44.7M [00:00<00:00, 90.5MB/s]
-100%|##########| 44.7M/44.7M [00:00<00:00, 102MB/s]
+ 27%|##7 | 12.2M/44.7M [00:00<00:00, 128MB/s]
+ 55%|#####4 | 24.4M/44.7M [00:00<00:00, 101MB/s]
+ 80%|#######9 | 35.7M/44.7M [00:00<00:00, 107MB/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 aaa8ac5659..8490280628 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 21.794 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 21.545 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 a7d2319d8f..f83dee6ccd 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>06:24.124</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>06:19.061</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>01:21.794</p></td>
+<td><p>01:21.545</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>01:17.727</p></td>
+<td><p>01:14.131</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>00:53.034</p></td>
+<td><p>00:52.717</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>00:35.504</p></td>
+<td><p>00:35.828</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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>00:30.906</p></td>
+<td><p>00:30.544</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>00:29.665</p></td>
+<td><p>00:30.531</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>00:26.794</p></td>
+<td><p>00:26.388</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>00:24.266</p></td>
+<td><p>00:24.511</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>00:21.723</p></td>
+<td><p>00:20.244</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>00:02.710</p></td>
+<td><p>00:02.622</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 5ffe745f9a..cf7e9611ea 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno.html
@@ -920,9 +920,10 @@ Top5 predictions:
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2548.1997 2547.6975 2550.0640 2546.8559 1.1696
+ 3341.0008 3340.7200 3343.2723 3338.7640 1.4712
</pre></div>
</div>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 3.163 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-model-on-adreno-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/2387d8448da213eb625e6b3d916327d4/deploy_model_on_adreno.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_model_on_adreno.py</span></code></a></p>
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 07791c6959..be4f07e68e 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)
- 16.1029 16.0874 16.2248 16.0180 0.0721
+ 16.6069 16.7762 17.1110 16.0513 0.3718
</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 55d0d21726..e79d6ac36d 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 +454,23 @@ 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]
- 5%|4 | 7.99M/170M [00:00<00:02, 74.3MB/s]
- 9%|9 | 16.0M/170M [00:00<00:02, 71.5MB/s]
- 14%|#4 | 24.0M/170M [00:00<00:02, 70.8MB/s]
- 19%|#8 | 32.0M/170M [00:00<00:01, 73.1MB/s]
- 24%|##3 | 40.0M/170M [00:00<00:01, 73.0MB/s]
- 28%|##8 | 48.0M/170M [00:00<00:01, 73.8MB/s]
- 32%|###2 | 55.0M/170M [00:00<00:01, 69.8MB/s]
- 37%|###6 | 62.3M/170M [00:00<00:01, 62.1MB/s]
- 42%|####2 | 72.0M/170M [00:01<00:01, 62.6MB/s]
- 47%|####7 | 80.0M/170M [00:01<00:01, 67.1MB/s]
- 52%|#####1 | 88.0M/170M [00:01<00:01, 69.5MB/s]
- 58%|#####8 | 98.7M/170M [00:01<00:00, 80.7MB/s]
- 63%|######2 | 107M/170M [00:01<00:00, 81.0MB/s]
- 69%|######9 | 118M/170M [00:01<00:00, 90.3MB/s]
- 74%|#######4 | 126M/170M [00:01<00:00, 79.8MB/s]
- 79%|#######9 | 134M/170M [00:01<00:00, 73.5MB/s]
- 83%|########3 | 142M/170M [00:02<00:00, 71.3MB/s]
- 88%|########8 | 150M/170M [00:02<00:00, 74.6MB/s]
- 93%|#########2| 158M/170M [00:02<00:00, 54.4MB/s]
- 98%|#########7| 166M/170M [00:02<00:00, 62.3MB/s]
-100%|##########| 170M/170M [00:02<00:00, 67.3MB/s]
+ 5%|4 | 7.99M/170M [00:00<00:02, 83.3MB/s]
+ 13%|#2 | 21.3M/170M [00:00<00:01, 116MB/s]
+ 19%|#9 | 32.3M/170M [00:00<00:01, 106MB/s]
+ 25%|##5 | 42.6M/170M [00:00<00:01, 105MB/s]
+ 31%|### | 52.6M/170M [00:00<00:01, 103MB/s]
+ 37%|###6 | 62.5M/170M [00:00<00:01, 102MB/s]
+ 43%|####2 | 72.3M/170M [00:00<00:01, 101MB/s]
+ 48%|####8 | 81.9M/170M [00:00<00:00, 96.7MB/s]
+ 55%|#####4 | 92.9M/170M [00:00<00:00, 102MB/s]
+ 60%|###### | 103M/170M [00:01<00:00, 99.3MB/s]
+ 66%|######6 | 112M/170M [00:01<00:00, 100MB/s]
+ 72%|#######2 | 123M/170M [00:01<00:00, 102MB/s]
+ 78%|#######7 | 132M/170M [00:01<00:00, 96.5MB/s]
+ 84%|########4 | 143M/170M [00:01<00:00, 102MB/s]
+ 90%|######### | 153M/170M [00:01<00:00, 88.6MB/s]
+ 99%|#########8| 167M/170M [00:01<00:00, 105MB/s]
+100%|##########| 170M/170M [00:01<00:00, 101MB/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 +568,7 @@ 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 34.456 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 27.786 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 12a8522a71..8c97dcad5f 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, 38.7MB/s]
-100%|##########| 13.6M/13.6M [00:00<00:00, 58.5MB/s]
+100%|#########9| 13.5M/13.6M [00:00<00:00, 142MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 141MB/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)
- 90.3979 90.2656 93.2883 90.1057 0.4102
+ 90.2892 90.2321 90.9815 90.0854 0.1888
</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 15.492 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 14.094 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 43f9607bb5..b1d7036796 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)
- 120.9549 120.9322 124.1480 120.2381 0.5169
+ 120.8368 120.7806 123.1834 120.0746 0.4294
</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 29.988 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 34.550 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 ac9a4be03d..caa56d99bd 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 41.583 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 20.755 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 eae966de22..26f96c9351 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -463,26 +463,23 @@ 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]
- 3%|2 | 3476/132723 [00:00<00:03, 34756.72KB/s]
- 8%|8 | 10875/132723 [00:00<00:02, 57827.60KB/s]
- 13%|#2 | 16658/132723 [00:00<00:04, 27190.02KB/s]
- 16%|#6 | 21691/132723 [00:00<00:03, 30515.28KB/s]
- 23%|##2 | 29999/132723 [00:00<00:02, 43065.12KB/s]
- 27%|##6 | 35488/132723 [00:00<00:02, 40446.62KB/s]
- 30%|### | 40317/132723 [00:01<00:02, 40208.61KB/s]
- 37%|###6 | 48976/132723 [00:01<00:01, 51664.61KB/s]
- 41%|####1 | 54831/132723 [00:01<00:01, 46908.56KB/s]
- 46%|####5 | 60726/132723 [00:01<00:01, 49868.54KB/s]
- 50%|####9 | 66151/132723 [00:01<00:01, 49220.32KB/s]
- 56%|#####6 | 74670/132723 [00:01<00:00, 58717.54KB/s]
- 62%|######1 | 81905/132723 [00:01<00:00, 55924.82KB/s]
- 68%|######8 | 90426/132723 [00:01<00:00, 63522.73KB/s]
- 74%|#######4 | 98288/132723 [00:01<00:00, 62993.22KB/s]
- 80%|######## | 106665/132723 [00:02<00:00, 68490.96KB/s]
- 87%|########6 | 115234/132723 [00:02<00:00, 73208.73KB/s]
- 93%|#########3| 123883/132723 [00:02<00:00, 76946.13KB/s]
-100%|#########9| 132591/132723 [00:02<00:00, 79852.22KB/s]
-100%|##########| 132723/132723 [00:02<00:00, 55436.23KB/s]
+ 5%|4 | 6147/132723 [00:00<00:02, 61464.05KB/s]
+ 11%|# | 14511/132723 [00:00<00:01, 74505.85KB/s]
+ 17%|#6 | 21962/132723 [00:00<00:01, 59073.28KB/s]
+ 21%|##1 | 28175/132723 [00:00<00:01, 59562.02KB/s]
+ 27%|##6 | 35273/132723 [00:00<00:01, 63275.70KB/s]
+ 33%|###3 | 44041/132723 [00:00<00:01, 71017.19KB/s]
+ 40%|###9 | 52694/132723 [00:00<00:01, 75854.81KB/s]
+ 46%|####6 | 61385/132723 [00:00<00:00, 79259.51KB/s]
+ 53%|#####2 | 70096/132723 [00:00<00:00, 81651.16KB/s]
+ 59%|#####9 | 78933/132723 [00:01<00:00, 83691.51KB/s]
+ 66%|######5 | 87584/132723 [00:01<00:00, 84539.97KB/s]
+ 73%|#######2 | 96432/132723 [00:01<00:00, 85722.14KB/s]
+ 79%|#######9 | 105229/132723 [00:01<00:00, 86395.14KB/s]
+ 86%|########5 | 114055/132723 [00:01<00:00, 86954.14KB/s]
+ 93%|#########2| 122821/132723 [00:01<00:00, 87162.81KB/s]
+ 99%|#########9| 131635/132723 [00:01<00:00, 87454.03KB/s]
+100%|##########| 132723/132723 [00:01<00:00, 79248.71KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -521,7 +518,7 @@ 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 40.281 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 34.842 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 f056aaf398..f2ca1a3747 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>15:12.314</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>14:51.380</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_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>03:40.281</p></td>
+<td><p>03:34.842</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>03:34.456</p></td>
+<td><p>03:27.786</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>02:29.988</p></td>
+<td><p>02:34.550</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>01:41.583</p></td>
+<td><p>01:20.755</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>01:15.492</p></td>
+<td><p>01:14.094</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>00:54.292</p></td>
+<td><p>01:03.163</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>00:41.392</p></td>
+<td><p>00:40.880</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>00:27.598</p></td>
+<td><p>00:27.715</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>00:27.226</p></td>
+<td><p>00:27.588</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 9f5daadd61..ca4803f18c 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.zip2840830e-0a34-445f-b9f6-dc7e4a1cd4ed 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.zip2e21da86-954a-4776-ad44-0a8bcc9cb300 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 16783efd2d..73c16ce056 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>00:53.794</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:53.112</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>00:49.939</p></td>
+<td><p>00:49.323</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>00:02.758</p></td>
+<td><p>00:02.708</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>00:01.090</p></td>
+<td><p>00:01.074</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 d043eac25b..6aa58b91a7 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: 21272us [21272us] (48.68%; 48.68%)
-FoldScaleAxis: 22428us [7us] (51.32%; 51.32%)
- FoldConstant: 22420us [1736us] (51.30%; 99.97%)
- InferType: 20684us [20684us] (47.33%; 92.26%)
+InferType: 22115us [22115us] (49.77%; 49.77%)
+FoldScaleAxis: 22321us [8us] (50.23%; 50.23%)
+ FoldConstant: 22313us [1754us] (50.21%; 99.97%)
+ InferType: 20559us [20559us] (46.27%; 92.14%)
</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: 20703us [20703us] (48.25%; 48.25%)
-FoldScaleAxis: 22200us [5us] (51.75%; 51.75%)
- FoldConstant: 22195us [1713us] (51.73%; 99.98%)
- InferType: 20482us [20482us] (47.74%; 92.28%)
+InferType: 20728us [20728us] (48.25%; 48.25%)
+FoldScaleAxis: 22232us [6us] (51.75%; 51.75%)
+ FoldConstant: 22227us [1738us] (51.74%; 99.98%)
+ InferType: 20489us [20489us] (47.69%; 92.18%)
</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 8a90266195..5a0f13b0a8 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: 54.141120 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 36.795871 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 cd875ca908..5d34b03ca3 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -867,7 +867,7 @@ 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: 7.909705 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 7.718617 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 388fcc08b1..0c77cf3165 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: 0.019201
-Baseline: 3.457197
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018811
+Baseline: 3.396133
</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: 0.305233
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.302338
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -589,7 +589,7 @@ 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: 0.338993
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.332977
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -644,7 +644,7 @@ 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: 0.118452
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.116724
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -721,7 +721,7 @@ 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: 0.109763
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109741
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -799,7 +799,7 @@ write to C when all the block results are ready.</p>
<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: 0.111094
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.110908
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -879,7 +879,7 @@ class Module:
<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: 0.148265
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.146963
</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 8928d242f4..dd9b95c5c3 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>00:35.477</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:34.845</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>00:32.722</p></td>
+<td><p>00:32.301</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>00:01.552</p></td>
+<td><p>00:01.475</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>00:01.203</p></td>
+<td><p>00:01.068</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 4e099c3822..9a714594a6 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>09:23.380</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>09:23.537</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>05:38.539</p></td>
+<td><p>05:42.565</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>01:40.846</p></td>
+<td><p>01:39.693</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>01:06.966</p></td>
+<td><p>01:05.839</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>00:29.196</p></td>
+<td><p>00:28.284</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>00:14.591</p></td>
+<td><p>00:14.117</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>00:13.241</p></td>
+<td><p>00:13.038</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 e247e043a1..02a19c8060 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
@@ -506,13 +506,13 @@ class Module:
def main(data: T.Buffer((1, 512, 7, 7), "float32"), kernel: T.Buffer((512, 512, 3, 3), "float32"), bias: T.Buffer((1, 512, 1, 1), "float32"), compute: T.Buffer((1, 512, 7, 7), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
blockIdx_x = T.env_thread("blockIdx.x")
- T.launch_thread(blockIdx_x, 8)
- conv2d_nchw = T.allocate([28], "float32", "local")
- pad_temp_shared = T.allocate([1296], "float32", "shared")
- kernel_shared = T.allocate([9216], "float32", "shared")
+ T.launch_thread(blockIdx_x, 112)
+ conv2d_nchw = T.allocate([7], "float32", "local")
+ pad_temp_shared = T.allocate([432], "float32", "shared")
+ kernel_shared = T.allocate([4608], "float32", "shared")
threadIdx_x = T.env_thread("threadIdx.x")
- T.launch_thread(threadIdx_x, 112)
- conv2d_nchw_1 = T.Buffer((28,), data=conv2d_nchw, scope="local")
+ T.launch_thread(threadIdx_x, 32)
+ conv2d_nchw_1 = T.Buffer((7,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
conv2d_nchw_1[1] = T.float32(0)
conv2d_nchw_1[2] = T.float32(0)
@@ -520,457 +520,786 @@ class Module:
conv2d_nchw_1[4] = T.float32(0)
conv2d_nchw_1[5] = T.float32(0)
conv2d_nchw_1[6] = T.float32(0)
- conv2d_nchw_1[7] = T.float32(0)
- conv2d_nchw_1[8] = T.float32(0)
- conv2d_nchw_1[9] = T.float32(0)
- conv2d_nchw_1[10] = T.float32(0)
- conv2d_nchw_1[11] = T.float32(0)
- conv2d_nchw_1[12] = T.float32(0)
- conv2d_nchw_1[13] = T.float32(0)
- conv2d_nchw_1[14] = T.float32(0)
- conv2d_nchw_1[15] = T.float32(0)
- conv2d_nchw_1[16] = T.float32(0)
- conv2d_nchw_1[17] = T.float32(0)
- conv2d_nchw_1[18] = T.float32(0)
- conv2d_nchw_1[19] = T.float32(0)
- conv2d_nchw_1[20] = T.float32(0)
- conv2d_nchw_1[21] = T.float32(0)
- conv2d_nchw_1[22] = T.float32(0)
- conv2d_nchw_1[23] = T.float32(0)
- conv2d_nchw_1[24] = T.float32(0)
- conv2d_nchw_1[25] = T.float32(0)
- conv2d_nchw_1[26] = T.float32(0)
- conv2d_nchw_1[27] = T.float32(0)
for rc_outer_outer in range(32):
cse_var_2: T.int32 = rc_outer_outer * 784
cse_var_1: T.int32 = rc_outer_outer * 144
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((1296,), data=pad_temp_shared, scope="shared")
+ pad_temp_shared_1 = T.Buffer((432,), data=pad_temp_shared, scope="shared")
data_1 = T.Buffer((25088,), data=data.data)
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(9 <= threadIdx_x_1 % 81 and threadIdx_x_1 % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 81 * 49 + threadIdx_x_1 % 81 // 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(9 <= (threadIdx_x_1 + 31) % 81 and (threadIdx_x_1 + 31) % 81 < 72 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 112) // 81 * 49 + (threadIdx_x_1 + 31) % 81 // 9 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(9 <= (threadIdx_x_1 + 62) % 81 and (threadIdx_x_1 + 62) % 81 < 72 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 81 * 49 + (threadIdx_x_1 + 62) % 81 // 9 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(9 <= (threadIdx_x_1 + 12) % 81 and (threadIdx_x_1 + 12) % 81 < 72 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 336) // 81 * 49 + (threadIdx_x_1 + 12) % 81 // 9 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(9 <= (threadIdx_x_1 + 43) % 81 and (threadIdx_x_1 + 43) % 81 < 72 and 1 <= (threadIdx_x_1 + 7) % 9 and (threadIdx_x_1 + 7) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 448) // 81 * 49 + (threadIdx_x_1 + 43) % 81 // 9 * 7 + (threadIdx_x_1 + 7) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(9 <= (threadIdx_x_1 + 74) % 81 and (threadIdx_x_1 + 74) % 81 < 72 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 560) // 81 * 49 + (threadIdx_x_1 + 74) % 81 // 9 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(9 <= (threadIdx_x_1 + 24) % 81 and (threadIdx_x_1 + 24) % 81 < 72 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 672) // 81 * 49 + (threadIdx_x_1 + 24) % 81 // 9 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(9 <= (threadIdx_x_1 + 55) % 81 and (threadIdx_x_1 + 55) % 81 < 72 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 784) // 81 * 49 + (threadIdx_x_1 + 55) % 81 // 9 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(9 <= (threadIdx_x_1 + 5) % 81 and (threadIdx_x_1 + 5) % 81 < 72 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 896) // 81 * 49 + (threadIdx_x_1 + 5) % 81 // 9 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(1 <= (threadIdx_x_1 // 9 + 4) % 9 and (threadIdx_x_1 + 36) % 81 < 72 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1008) // 81 * 49 + (threadIdx_x_1 // 9 + 4) % 9 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(9 <= (threadIdx_x_1 + 67) % 81 and (threadIdx_x_1 + 67) % 81 < 72 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1120) // 81 * 49 + (threadIdx_x_1 + 67) % 81 // 9 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 112):
- if T.likely(threadIdx_x_1 < 64):
- pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(threadIdx_x_1 < 55 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1232) // 81 * 49 + (threadIdx_x_1 + 17) % 81 // 9 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 and threadIdx_x_1 % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + threadIdx_x_1 // 27 * 49 + threadIdx_x_1 % 27 // 9 * 7 + blockIdx_x % 7 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 32] = T.if_then_else(1 <= (threadIdx_x_1 + 5) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 5) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 32) // 27 * 49 + (threadIdx_x_1 + 5) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 64] = T.if_then_else(1 <= (threadIdx_x_1 + 10) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 10) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 64) // 27 * 49 + (threadIdx_x_1 + 10) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 96] = T.if_then_else(1 <= (threadIdx_x_1 + 15) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 15) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 96) // 27 * 49 + (threadIdx_x_1 + 15) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 128] = T.if_then_else(1 <= (threadIdx_x_1 + 20) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 20) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 128) // 27 * 49 + (threadIdx_x_1 + 20) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 160] = T.if_then_else(1 <= (threadIdx_x_1 + 25) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 25) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 7) % 9 and (threadIdx_x_1 + 7) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 160) // 27 * 49 + (threadIdx_x_1 + 25) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 7) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 192] = T.if_then_else(1 <= (threadIdx_x_1 + 3) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 3) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 3) % 9 and (threadIdx_x_1 + 3) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 192) // 27 * 49 + (threadIdx_x_1 + 3) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 3) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= (threadIdx_x_1 + 8) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 8) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 8) % 9 and (threadIdx_x_1 + 8) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 27 * 49 + (threadIdx_x_1 + 8) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 8) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 256] = T.if_then_else(1 <= (threadIdx_x_1 + 13) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 13) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 4) % 9 and (threadIdx_x_1 + 4) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 256) // 27 * 49 + (threadIdx_x_1 + 13) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 4) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 288] = T.if_then_else(1 <= blockIdx_x % 7 + (threadIdx_x_1 // 9 + 2) % 3 and blockIdx_x % 7 + (threadIdx_x_1 // 9 + 2) % 3 < 8 and 1 <= threadIdx_x_1 % 9 and threadIdx_x_1 % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 288) // 27 * 49 + blockIdx_x % 7 * 7 + (threadIdx_x_1 // 9 + 2) % 3 * 7 + threadIdx_x_1 % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 320] = T.if_then_else(1 <= (threadIdx_x_1 + 23) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 23) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 5) % 9 and (threadIdx_x_1 + 5) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 320) // 27 * 49 + (threadIdx_x_1 + 23) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 5) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 352] = T.if_then_else(1 <= (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 1) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 1) % 9 and (threadIdx_x_1 + 1) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 352) // 27 * 49 + (threadIdx_x_1 + 1) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 1) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ pad_temp_shared_1[threadIdx_x_1 + 384] = T.if_then_else(1 <= (threadIdx_x_1 + 6) % 27 // 9 + blockIdx_x % 7 and (threadIdx_x_1 + 6) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 6) % 9 and (threadIdx_x_1 + 6) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 384) // 27 * 49 + (threadIdx_x_1 + 6) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 6) % 9 - 8], T.float32(0))
+ with T.launch_thread(threadIdx_x_1, 32):
+ if T.likely(threadIdx_x_1 < 16):
+ pad_temp_shared_1[threadIdx_x_1 + 416] = T.if_then_else((threadIdx_x_1 + 11) % 27 // 9 + blockIdx_x % 7 < 8 and 1 <= (threadIdx_x_1 + 2) % 9 and (threadIdx_x_1 + 2) % 9 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 416) // 27 * 49 + (threadIdx_x_1 + 11) % 27 // 9 * 7 + blockIdx_x % 7 * 7 + (threadIdx_x_1 + 2) % 9 - 8], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
- kernel_shared_1 = T.Buffer((9216,), data=kernel_shared, scope="shared")
+ kernel_shared_1 = T.Buffer((4608,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 112) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 336) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 560) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 672) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 784) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 896) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 32256]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1120) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1232) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1344) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1456) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1568) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1680) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1792) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 1904) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 64512]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2240) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2352] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2464) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2576] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2576) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2688) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2800] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2800) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 2912) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3024] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 96768]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3136] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3136) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3248] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3248) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3360] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3360) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3472] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3472) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3584] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3584) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3696] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3696) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3808] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3808) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 3920] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 3920) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4032] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 129024]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4144] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4144) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4256] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4368] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4368) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4480] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4592] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4592) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4704] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4704) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4816] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4816) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 4928] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 4928) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5040] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 161280]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5152] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5152) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5264] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5264) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5376] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5376) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5488] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5488) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5600] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5600) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5712] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5712) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5824] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5824) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 5936] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 5936) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6048] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 193536]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6160] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6272] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6272) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6384] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6496] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6496) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6608] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6608) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6720] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6720) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6832] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6832) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 6944] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 6944) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7056] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 225792]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7168] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7168) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7280] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7280) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7392] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7392) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7504] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7504) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7616] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7616) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7728] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7728) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7840] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7840) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 7952] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 7952) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8064] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 258048]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8176] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8176) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8288] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8288) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8400] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8400) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8512] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8624] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8624) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8736] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8736) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) % 48 * 3 + threadIdx_x_2 % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8848] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8848) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 8960] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 8960) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
- with T.launch_thread(threadIdx_x_2, 112):
- kernel_shared_1[threadIdx_x_2 + 9072] = kernel_1[blockIdx_x * 294912 + cse_var_1 + threadIdx_x_2 + 290304]
- with T.launch_thread(threadIdx_x_2, 112):
- if T.likely(threadIdx_x_2 < 32):
- kernel_shared_1[threadIdx_x_2 + 9184] = kernel_1[blockIdx_x * 294912 + (threadIdx_x_2 + 9184) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
- for rc_outer_inner, ry_outer_inner, rx_outer_inner in T.grid(2, 3, 3):
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 144]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 288]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 1] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 2] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 3] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 4] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 5] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 6] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 432]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 9]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 153]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 297]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 81] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 82] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 83] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 84] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 85] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 86] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 87] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 441]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 18]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 162]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 306]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 162] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 163] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 164] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 165] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 166] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 167] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 168] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 450]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 27]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 171]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 315]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 243] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 244] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 245] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 246] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 247] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 248] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 249] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 459]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 36]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 180]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 324]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 324] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 325] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 326] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 327] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 328] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 329] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 330] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 468]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 45]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 189]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 333]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 405] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 406] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 407] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 408] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 409] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 410] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 411] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 477]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 54]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 198]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 342]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 486] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 487] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 488] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 489] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 490] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 491] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 492] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 486]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 63]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 207]
- conv2d_nchw_1[14] = conv2d_nchw_1[14] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[15] = conv2d_nchw_1[15] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[16] = conv2d_nchw_1[16] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[17] = conv2d_nchw_1[17] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[18] = conv2d_nchw_1[18] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[19] = conv2d_nchw_1[19] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[20] = conv2d_nchw_1[20] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 351]
- conv2d_nchw_1[21] = conv2d_nchw_1[21] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 567] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[22] = conv2d_nchw_1[22] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 568] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[23] = conv2d_nchw_1[23] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 569] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[24] = conv2d_nchw_1[24] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 570] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[25] = conv2d_nchw_1[25] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 571] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[26] = conv2d_nchw_1[26] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 572] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- conv2d_nchw_1[27] = conv2d_nchw_1[27] + pad_temp_shared_1[rc_outer_inner * 648 + ry_outer_inner * 9 + threadIdx_x % 7 * 9 + rx_outer_inner + 573] * kernel_shared_1[threadIdx_x // 7 * 576 + rc_outer_inner * 72 + ry_outer_inner * 3 + rx_outer_inner + 495]
- for i1_inner, i3_inner in T.grid(4, 7):
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 32] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + (threadIdx_x_2 + 32) // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 64] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + (threadIdx_x_2 + 64) // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 96] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 96]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 128] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 160] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 192] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 192) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 256] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 288] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 9216]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 320] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 320) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 352] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 384] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 416] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 416) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 480] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 512] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 544] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 544) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 576] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 18432]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 608] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 608) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 640] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 640) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 672) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 704] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 704) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 736] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 736) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 768] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 768) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 800] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 800) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 832] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 832) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 864] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 27648]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 896) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 928] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 928) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 960] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 960) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 992] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 992) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1024] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1024) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1056] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1056) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1088] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1088) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1120) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1152] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 36864]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1184] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1184) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1216] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1216) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1248] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1248) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1280] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1280) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1312] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1312) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1344) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1376] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1376) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1408] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1408) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1440] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 46080]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1472] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1472) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1504] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1504) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1536] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1536) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1568) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1600] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1600) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1632] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1632) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1664] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1664) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1696] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1696) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1728] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 55296]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1760] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1760) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1792) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1824] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1824) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1856] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1856) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1888] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1888) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1920] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1920) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1952] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1952) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 1984] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 1984) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 64512]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2048] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2048) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2080] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2080) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2112] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2112) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2144] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2144) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2176] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2176) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2208] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2208) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2240) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2272] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2272) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2304] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 73728]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2336] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2336) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2368] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2368) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2400] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2400) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2432] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2432) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2464) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2496] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2496) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2528] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2528) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2560] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2560) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2592] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 82944]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2624] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2624) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2656] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2656) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2688) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2720] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2720) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2752] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2752) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2784] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2784) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2816] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2816) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2848] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2848) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2880] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 92160]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2912) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2944] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2944) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 2976] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 2976) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3008] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3008) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3040] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3040) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3072] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3072) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3104] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3104) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3136] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3136) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3168] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 101376]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3200] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3200) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3232] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3232) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3264] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3264) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3296] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3296) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3328] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3328) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3360] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3360) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3392] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3392) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3424] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3424) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3456] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 110592]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3488] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3488) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3520] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3520) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3552] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3552) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3584] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3584) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3616] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3616) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3648] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3648) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3680] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3680) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3712] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3712) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3744] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 119808]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3776] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3776) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3808] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3808) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3840] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3840) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3872] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3872) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3904] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3904) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3936] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3936) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 3968] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 3968) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4000] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4000) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4032] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 129024]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4064] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4064) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4096] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4096) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4128] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4128) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4160] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4160) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4192] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4192) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4224] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4224) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4256] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4256) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4288] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4288) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4320] = kernel_1[blockIdx_x // 7 * 147456 + cse_var_1 + threadIdx_x_2 + 138240]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4352] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4352) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 32) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4384] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4384) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 64) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4416] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4416) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 32) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4448] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4448) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 128) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4480] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4480) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 16) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4512] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4512) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 // 3 + 16) * 3 + threadIdx_x_2 % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4544] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4544) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 80) % 144 // 3 * 3 + (threadIdx_x_2 + 2) % 3]
+ with T.launch_thread(threadIdx_x_2, 32):
+ kernel_shared_1[threadIdx_x_2 + 4576] = kernel_1[blockIdx_x // 7 * 147456 + (threadIdx_x_2 + 4576) // 144 * 4608 + cse_var_1 + (threadIdx_x_2 + 112) % 144 // 3 * 3 + (threadIdx_x_2 + 1) % 3]
+ for ry_outer_inner in range(3):
+ cse_var_115: T.int32 = ry_outer_inner * 9
+ cse_var_114: T.int32 = cse_var_115 + 88
+ cse_var_113: T.int32 = cse_var_115 + 87
+ cse_var_112: T.int32 = cse_var_115 + 86
+ cse_var_111: T.int32 = cse_var_115 + 85
+ cse_var_110: T.int32 = cse_var_115 + 84
+ cse_var_109: T.int32 = cse_var_115 + 83
+ cse_var_108: T.int32 = cse_var_115 + 82
+ cse_var_107: T.int32 = cse_var_115 + 7
+ cse_var_106: T.int32 = cse_var_115 + 61
+ cse_var_105: T.int32 = cse_var_115 + 60
+ cse_var_104: T.int32 = cse_var_115 + 6
+ cse_var_103: T.int32 = cse_var_115 + 59
+ cse_var_102: T.int32 = cse_var_115 + 58
+ cse_var_101: T.int32 = cse_var_115 + 57
+ cse_var_100: T.int32 = cse_var_115 + 56
+ cse_var_99: T.int32 = cse_var_115 + 55
+ cse_var_98: T.int32 = cse_var_115 + 5
+ cse_var_97: T.int32 = cse_var_115 + 412
+ cse_var_96: T.int32 = cse_var_115 + 411
+ cse_var_95: T.int32 = cse_var_115 + 410
+ cse_var_94: T.int32 = cse_var_115 + 409
+ cse_var_93: T.int32 = cse_var_115 + 408
+ cse_var_92: T.int32 = cse_var_115 + 407
+ cse_var_91: T.int32 = cse_var_115 + 406
+ cse_var_90: T.int32 = cse_var_115 + 4
+ cse_var_89: T.int32 = cse_var_115 + 385
+ cse_var_88: T.int32 = cse_var_115 + 384
+ cse_var_87: T.int32 = cse_var_115 + 383
+ cse_var_86: T.int32 = cse_var_115 + 382
+ cse_var_85: T.int32 = cse_var_115 + 381
+ cse_var_84: T.int32 = cse_var_115 + 380
+ cse_var_83: T.int32 = cse_var_115 + 379
+ cse_var_82: T.int32 = cse_var_115 + 358
+ cse_var_81: T.int32 = cse_var_115 + 357
+ cse_var_80: T.int32 = cse_var_115 + 356
+ cse_var_79: T.int32 = cse_var_115 + 355
+ cse_var_78: T.int32 = cse_var_115 + 354
+ cse_var_77: T.int32 = cse_var_115 + 353
+ cse_var_76: T.int32 = cse_var_115 + 352
+ cse_var_75: T.int32 = cse_var_115 + 34
+ cse_var_74: T.int32 = cse_var_115 + 331
+ cse_var_73: T.int32 = cse_var_115 + 330
+ cse_var_72: T.int32 = cse_var_115 + 33
+ cse_var_71: T.int32 = cse_var_115 + 329
+ cse_var_70: T.int32 = cse_var_115 + 328
+ cse_var_69: T.int32 = cse_var_115 + 327
+ cse_var_68: T.int32 = cse_var_115 + 326
+ cse_var_67: T.int32 = cse_var_115 + 325
+ cse_var_66: T.int32 = cse_var_115 + 32
+ cse_var_65: T.int32 = cse_var_115 + 31
+ cse_var_64: T.int32 = cse_var_115 + 304
+ cse_var_63: T.int32 = cse_var_115 + 303
+ cse_var_62: T.int32 = cse_var_115 + 302
+ cse_var_61: T.int32 = cse_var_115 + 301
+ cse_var_60: T.int32 = cse_var_115 + 300
+ cse_var_59: T.int32 = cse_var_115 + 30
+ cse_var_58: T.int32 = cse_var_115 + 3
+ cse_var_57: T.int32 = cse_var_115 + 299
+ cse_var_56: T.int32 = cse_var_115 + 298
+ cse_var_55: T.int32 = cse_var_115 + 29
+ cse_var_54: T.int32 = cse_var_115 + 28
+ cse_var_53: T.int32 = cse_var_115 + 277
+ cse_var_52: T.int32 = cse_var_115 + 276
+ cse_var_51: T.int32 = cse_var_115 + 275
+ cse_var_50: T.int32 = cse_var_115 + 274
+ cse_var_49: T.int32 = cse_var_115 + 273
+ cse_var_48: T.int32 = cse_var_115 + 272
+ cse_var_47: T.int32 = cse_var_115 + 271
+ cse_var_46: T.int32 = cse_var_115 + 250
+ cse_var_45: T.int32 = cse_var_115 + 249
+ cse_var_44: T.int32 = cse_var_115 + 248
+ cse_var_43: T.int32 = cse_var_115 + 247
+ cse_var_42: T.int32 = cse_var_115 + 246
+ cse_var_41: T.int32 = cse_var_115 + 245
+ cse_var_40: T.int32 = cse_var_115 + 244
+ cse_var_39: T.int32 = cse_var_115 + 223
+ cse_var_38: T.int32 = cse_var_115 + 222
+ cse_var_37: T.int32 = cse_var_115 + 221
+ cse_var_36: T.int32 = cse_var_115 + 220
+ cse_var_35: T.int32 = cse_var_115 + 219
+ cse_var_34: T.int32 = cse_var_115 + 218
+ cse_var_33: T.int32 = cse_var_115 + 217
+ cse_var_32: T.int32 = cse_var_115 + 2
+ cse_var_31: T.int32 = cse_var_115 + 196
+ cse_var_30: T.int32 = cse_var_115 + 195
+ cse_var_29: T.int32 = cse_var_115 + 194
+ cse_var_28: T.int32 = cse_var_115 + 193
+ cse_var_27: T.int32 = cse_var_115 + 192
+ cse_var_26: T.int32 = cse_var_115 + 191
+ cse_var_25: T.int32 = cse_var_115 + 190
+ cse_var_24: T.int32 = cse_var_115 + 169
+ cse_var_23: T.int32 = cse_var_115 + 168
+ cse_var_22: T.int32 = cse_var_115 + 167
+ cse_var_21: T.int32 = cse_var_115 + 166
+ cse_var_20: T.int32 = cse_var_115 + 165
+ cse_var_19: T.int32 = cse_var_115 + 164
+ cse_var_18: T.int32 = cse_var_115 + 163
+ cse_var_17: T.int32 = cse_var_115 + 142
+ cse_var_16: T.int32 = cse_var_115 + 141
+ cse_var_15: T.int32 = cse_var_115 + 140
+ cse_var_14: T.int32 = cse_var_115 + 139
+ cse_var_13: T.int32 = cse_var_115 + 138
+ cse_var_12: T.int32 = cse_var_115 + 137
+ cse_var_11: T.int32 = cse_var_115 + 136
+ cse_var_10: T.int32 = cse_var_115 + 115
+ cse_var_9: T.int32 = cse_var_115 + 114
+ cse_var_8: T.int32 = cse_var_115 + 113
+ cse_var_7: T.int32 = cse_var_115 + 112
+ cse_var_6: T.int32 = cse_var_115 + 111
+ cse_var_5: T.int32 = cse_var_115 + 110
+ cse_var_4: T.int32 = cse_var_115 + 109
+ cse_var_3: T.int32 = cse_var_115 + 1
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_3] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_99] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_4] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 135] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_11] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 162] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_18] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 189] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_25] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 216] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_33] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 243] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_40] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 270] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_47] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 297] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_56] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 324] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_67] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 351] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_76] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 378] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_83] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_115 + 405] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_91] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_3] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_54] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_99] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_108] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_4] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_11] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_18] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_25] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_33] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_40] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_47] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_56] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_67] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_76] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_83] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_91] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_32] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_55] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_100] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_109] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_5] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_12] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_19] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_26] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_34] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_41] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_48] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_57] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_68] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_77] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_84] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_92] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_58] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_59] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_101] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_110] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_6] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_13] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_20] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_27] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_42] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_49] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_60] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_69] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_78] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_85] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_93] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_90] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_65] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_102] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_111] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_7] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_14] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_21] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_28] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_36] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_43] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_50] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_61] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_70] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_79] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_86] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_94] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_98] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_107] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_66] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_75] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_103] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_106] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_112] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_114] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_10] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_15] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_17] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_22] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_24] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_29] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_31] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_37] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_39] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_44] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_46] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_51] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_53] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_64] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_71] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_74] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_80] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_82] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_87] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_95] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[cse_var_97] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_104] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_107] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 1]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 8] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 2]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_72] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 9]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_75] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 10]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 35] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 11]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_105] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 18]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_106] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 19]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 62] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 20]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_113] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 27]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_114] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 28]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 29]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_9] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 36]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_10] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 37]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 116] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 38]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_16] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 45]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_17] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 46]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 143] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 47]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_23] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 54]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_24] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 55]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 170] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 56]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_30] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 63]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_31] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 64]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 197] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 65]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_38] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 72]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_39] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 73]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 224] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 74]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_45] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 81]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_46] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 82]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 251] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 83]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_52] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 90]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_53] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 91]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 278] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 92]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_63] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 99]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_64] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 100]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 305] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 101]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_73] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 108]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_74] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 109]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 332] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 110]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_81] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 117]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_82] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 118]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 359] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 119]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_88] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 126]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_89] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 127]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 386] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 128]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_96] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 135]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_97] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 136]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[cse_var_115 + 413] * kernel_shared_1[threadIdx_x * 144 + ry_outer_inner * 3 + 137]
+ for i3_inner in range(7):
compute_1 = T.Buffer((25088,), data=compute.data)
bias_1 = T.Buffer((512,), data=bias.data)
- compute_1[blockIdx_x * 3136 + threadIdx_x // 7 * 196 + i1_inner * 49 + threadIdx_x % 7 * 7 + i3_inner] = T.max(conv2d_nchw_1[i1_inner * 7 + i3_inner] + bias_1[blockIdx_x * 64 + threadIdx_x // 7 * 4 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x // 7 * 1568 + threadIdx_x * 49 + blockIdx_x % 7 * 7 + i3_inner] = T.max(conv2d_nchw_1[i3_inner] + bias_1[blockIdx_x // 7 * 32 + threadIdx_x], T.float32(0))
</pre></div>
</div>
</div>
@@ -1004,7 +1333,7 @@ class Module:
<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: 0.398 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.251 ms
</pre></div>
</div>
</div>
@@ -1033,33 +1362,33 @@ conv2d_nchw_nn_o_i, conv2d_nchw_nn_i = s[conv2d_nchw].split(conv2d_nchw_nn, fact
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
-conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=4)
+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=16)
+conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=32)
conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
-conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
+conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
-conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=7)
-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_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=7)
conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
-conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=8)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=2)
+conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=16)
+conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=1)
conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=3)
-conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
-conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
+conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2d_nc [...]
compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=4)
-compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=16)
+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=32)
compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
-compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
+compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
@@ -1082,12 +1411,12 @@ 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=112)
+kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=32)
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)
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=112)
+pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=32)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
@@ -1107,10 +1436,10 @@ CUDA source code:
#define int64_t long long
#define uint64_t unsigned long long
#endif
-extern "C" __global__ void __launch_bounds__(112) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
- float conv2d_nchw[28];
- __shared__ float pad_temp_shared[1296];
- __shared__ float kernel_shared[9216];
+extern "C" __global__ void __launch_bounds__(32) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
+ float conv2d_nchw[7];
+ __shared__ float pad_temp_shared[432];
+ __shared__ float kernel_shared[4608];
conv2d_nchw[0] = 0.000000e+00f;
conv2d_nchw[1] = 0.000000e+00f;
conv2d_nchw[2] = 0.000000e+00f;
@@ -1118,364 +1447,510 @@ extern "C" __global__ void __launch_bounds__(112) default_function_ker
conv2d_nchw[4] = 0.000000e+00f;
conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[6] = 0.000000e+00f;
- conv2d_nchw[7] = 0.000000e+00f;
- conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[9] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[11] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
- conv2d_nchw[13] = 0.000000e+00f;
- conv2d_nchw[14] = 0.000000e+00f;
- conv2d_nchw[15] = 0.000000e+00f;
- conv2d_nchw[16] = 0.000000e+00f;
- conv2d_nchw[17] = 0.000000e+00f;
- conv2d_nchw[18] = 0.000000e+00f;
- conv2d_nchw[19] = 0.000000e+00f;
- conv2d_nchw[20] = 0.000000e+00f;
- conv2d_nchw[21] = 0.000000e+00f;
- conv2d_nchw[22] = 0.000000e+00f;
- conv2d_nchw[23] = 0.000000e+00f;
- conv2d_nchw[24] = 0.000000e+00f;
- conv2d_nchw[25] = 0.000000e+00f;
- conv2d_nchw[26] = 0.000000e+00f;
- conv2d_nchw[27] = 0.000000e+00f;
for (int rc_outer_outer = 0; rc_outer_outer < 32; ++rc_outer_outer) {
__syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = (((((9 <= (((int)threadIdx.x) % 81)) && ((((int)threadIdx.x) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + ((((int)threadIdx.x) / 81) * 49)) + (((((int)threadIdx.x) % 81) / 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((9 <= ((((int)threadIdx.x) + 31) % 81)) && (((((int)threadIdx.x) + 31) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 112) / 81) * 49)) + ((((((int)threadIdx.x) + 31) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((9 <= ((((int)threadIdx.x) + 62) % 81)) && (((((int)threadIdx.x) + 62) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 224) / 81) * 49)) + ((((((int)threadIdx.x) + 62) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((9 <= ((((int)threadIdx.x) + 12) % 81)) && (((((int)threadIdx.x) + 12) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 336) / 81) * 49)) + ((((((int)threadIdx.x) + 12) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 448)] = (((((9 <= ((((int)threadIdx.x) + 43) % 81)) && (((((int)threadIdx.x) + 43) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 448) / 81) * 49)) + ((((((int)threadIdx.x) + 43) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 7) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((9 <= ((((int)threadIdx.x) + 74) % 81)) && (((((int)threadIdx.x) + 74) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 560) / 81) * 49)) + ((((((int)threadIdx.x) + 74) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((9 <= ((((int)threadIdx.x) + 24) % 81)) && (((((int)threadIdx.x) + 24) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 672) / 81) * 49)) + ((((((int)threadIdx.x) + 24) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 6) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((9 <= ((((int)threadIdx.x) + 55) % 81)) && (((((int)threadIdx.x) + 55) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 784) / 81) * 49)) + ((((((int)threadIdx.x) + 55) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((9 <= ((((int)threadIdx.x) + 5) % 81)) && (((((int)threadIdx.x) + 5) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 896) / 81) * 49)) + ((((((int)threadIdx.x) + 5) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 5) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1008)] = (((((1 <= (((((int)threadIdx.x) / 9) + 4) % 9)) && (((((int)threadIdx.x) + 36) % 81) < 72)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1008) / 81) * 49)) + ((((((int)threadIdx.x) / 9) + 4) % 9) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((9 <= ((((int)threadIdx.x) + 67) % 81)) && (((((int)threadIdx.x) + 67) % 81) < 72)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1120) / 81) * 49)) + ((((((int)threadIdx.x) + 67) % 81) / 9) * 7)) + ((((int)threadIdx.x) + 4) % 9)) - 8)] : 0.000000e+00f);
- if (((int)threadIdx.x) < 64) {
- pad_temp_shared[(((int)threadIdx.x) + 1232)] = ((((((int)threadIdx.x) < 55) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[(((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 1232) / 81) * 49)) + (((((int)threadIdx.x) + 17) / 9) * 7)) + ((((int)threadIdx.x) + 8) % 9)) - 8)] : 0.000000e+00f);
- }
- kernel_shared[((int)threadIdx.x)] = kernel[(((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x))];
- kernel_shared[(((int)threadIdx.x) + 112)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 112) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 224)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 224) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 336) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 448) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 560) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 672) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 784) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 896) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 32256)];
- kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1120) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1232) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1344) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1456) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1568) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1680) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1792) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 1904) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 64512)];
- kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2240) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2352)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2352) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2464) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2576)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2576) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2688) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2800)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2800) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 2912) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3024)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96768)];
- kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3136) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3248)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3248) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3360)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3360) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3472)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3472) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3584) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3696)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3696) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3808)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3808) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 3920)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 3920) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 129024)];
- kernel_shared[(((int)threadIdx.x) + 4144)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4144) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4256)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4256) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4368)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4368) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4480) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4592)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4592) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4704)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4704) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4816)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4816) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 4928)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 4928) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5040)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 161280)];
- kernel_shared[(((int)threadIdx.x) + 5152)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5152) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5264)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5264) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5376)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5376) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5488)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5488) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5600)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5600) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5712)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5712) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5824)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5824) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 5936)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 5936) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6048)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 193536)];
- kernel_shared[(((int)threadIdx.x) + 6160)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6160) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6272)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6272) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6384)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6384) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6496)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6496) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6608)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6608) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6720)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6720) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6832)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6832) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 6944)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 6944) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7056)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 225792)];
- kernel_shared[(((int)threadIdx.x) + 7168)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7168) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7280)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7280) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7392)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7392) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7504)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7504) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7616)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7616) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7728)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7728) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7840)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7840) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 7952)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 7952) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8064)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 258048)];
- kernel_shared[(((int)threadIdx.x) + 8176)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8176) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 112) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8288)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8288) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 80) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8400)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8400) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 16) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8512)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8512) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8624)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8624) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8736)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8736) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) / 3) + 32) % 48) * 3)) + (((int)threadIdx.x) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8848)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8848) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 64) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
- kernel_shared[(((int)threadIdx.x) + 8960)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 8960) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
- kernel_shared[(((int)threadIdx.x) + 9072)] = kernel[((((((int)blockIdx.x) * 294912) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 290304)];
- if (((int)threadIdx.x) < 32) {
- kernel_shared[(((int)threadIdx.x) + 9184)] = kernel[(((((((int)blockIdx.x) * 294912) + (((((int)threadIdx.x) + 9184) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 32)] = (((((1 <= ((((((int)threadIdx.x) + 5) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 5) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 32) / 27) * 49)) + ((((((int)threadIdx.x) + 5) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)thr [...]
+ pad_temp_shared[(((int)threadIdx.x) + 64)] = (((((1 <= ((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 64) / 27) * 49)) + ((((((int)threadIdx.x) + 10) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int) [...]
+ pad_temp_shared[(((int)threadIdx.x) + 96)] = (((((1 <= ((((((int)threadIdx.x) + 15) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 15) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 96) / 27) * 49)) + ((((((int)threadIdx.x) + 15) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int) [...]
+ pad_temp_shared[(((int)threadIdx.x) + 128)] = (((((1 <= ((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 128) / 27) * 49)) + ((((((int)threadIdx.x) + 20) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((in [...]
+ pad_temp_shared[(((int)threadIdx.x) + 160)] = (((((1 <= ((((((int)threadIdx.x) + 25) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 25) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 7) % 9))) && (((((int)threadIdx.x) + 7) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 160) / 27) * 49)) + ((((((int)threadIdx.x) + 25) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((in [...]
+ pad_temp_shared[(((int)threadIdx.x) + 192)] = (((((1 <= ((((((int)threadIdx.x) + 3) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 3) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 192) / 27) * 49)) + ((((((int)threadIdx.x) + 3) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)t [...]
+ pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= ((((((int)threadIdx.x) + 8) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 8) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 8) % 9))) && (((((int)threadIdx.x) + 8) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 224) / 27) * 49)) + ((((((int)threadIdx.x) + 8) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)t [...]
+ pad_temp_shared[(((int)threadIdx.x) + 256)] = (((((1 <= ((((((int)threadIdx.x) + 13) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 13) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 4) % 9))) && (((((int)threadIdx.x) + 4) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 256) / 27) * 49)) + ((((((int)threadIdx.x) + 13) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((in [...]
+ pad_temp_shared[(((int)threadIdx.x) + 288)] = (((((1 <= ((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) / 9) + 2) % 3))) && (((((int)blockIdx.x) % 7) + (((((int)threadIdx.x) / 9) + 2) % 3)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 288) / 27) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + ((((((int)threadIdx.x) / 9) + 2) % 3) * 7)) + (((int)threadIdx.x) % 9) [...]
+ pad_temp_shared[(((int)threadIdx.x) + 320)] = (((((1 <= ((((((int)threadIdx.x) + 23) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 23) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 5) % 9))) && (((((int)threadIdx.x) + 5) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 320) / 27) * 49)) + ((((((int)threadIdx.x) + 23) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((in [...]
+ pad_temp_shared[(((int)threadIdx.x) + 352)] = (((((1 <= ((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 1) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 352) / 27) * 49)) + ((((((int)threadIdx.x) + 1) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)t [...]
+ pad_temp_shared[(((int)threadIdx.x) + 384)] = (((((1 <= ((((((int)threadIdx.x) + 6) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 6) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 6) % 9))) && (((((int)threadIdx.x) + 6) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 384) / 27) * 49)) + ((((((int)threadIdx.x) + 6) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)t [...]
+ if (((int)threadIdx.x) < 16) {
+ pad_temp_shared[(((int)threadIdx.x) + 416)] = (((((((((int)threadIdx.x) + 11) / 9) + (((int)blockIdx.x) % 7)) < 8) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 784) + (((((int)threadIdx.x) + 416) / 27) * 49)) + (((((int)threadIdx.x) + 11) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)] : 0.000000e+00f);
}
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x))];
+ kernel_shared[(((int)threadIdx.x) + 32)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 96)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 160)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 160) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 192) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 224) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 256) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 288)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 9216)];
+ kernel_shared[(((int)threadIdx.x) + 320)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 320) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 352)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 352) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 384) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 416)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 416) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 448) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 480)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 480) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 512) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 544)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 544) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 576)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 18432)];
+ kernel_shared[(((int)threadIdx.x) + 608)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 608) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 640) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 672) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 704)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 704) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 736)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 736) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 768) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 800)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 800) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 832)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 832) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 864)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 27648)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 896) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 928)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 928) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 960) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 992)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 992) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1024) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1056)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1056) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1088) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1120) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 36864)];
+ kernel_shared[(((int)threadIdx.x) + 1184)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1184) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1216) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1248)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1248) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1280) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1312)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1312) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1344) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1376)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1376) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1408) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1440)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 46080)];
+ kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1472) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1504)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1504) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1536) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1568) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1600) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1632)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1632) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1664) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1696)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1696) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 55296)];
+ kernel_shared[(((int)threadIdx.x) + 1760)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1760) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1792) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1824)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1824) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1856) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1888)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1888) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1920) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 1952)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1952) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 1984) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2048) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2080)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2080) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2112) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2144)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2144) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2176) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2208)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2208) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2240) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2272)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2272) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 73728)];
+ kernel_shared[(((int)threadIdx.x) + 2336)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2336) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2368) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2400)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2400) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2432) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2464) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2496) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2528)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2528) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2560) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2592)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 82944)];
+ kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2624) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2656)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2656) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2688) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2720)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2720) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2752) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2784)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2784) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2816) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2848)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2848) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 92160)];
+ kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2912) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2944) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 2976)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 2976) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3008) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3040)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3040) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3072)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3072) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3104)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3104) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3136) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3168)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 101376)];
+ kernel_shared[(((int)threadIdx.x) + 3200)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3200) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3232)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3232) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3264)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3264) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3296)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3296) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3328)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3328) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3360)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3360) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3392)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3392) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3424)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3424) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3456)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 110592)];
+ kernel_shared[(((int)threadIdx.x) + 3488)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3488) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3520)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3520) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3552)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3552) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3584) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3616)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3616) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3648)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3648) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3680)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3680) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3712)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3712) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3744)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 119808)];
+ kernel_shared[(((int)threadIdx.x) + 3776)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3776) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3808)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3808) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3840)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3840) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 3872)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3872) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3904)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3904) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 3936)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3936) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 3968)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 3968) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4000)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4000) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 129024)];
+ kernel_shared[(((int)threadIdx.x) + 4064)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4064) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4096)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4096) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4128)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4128) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 4160)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4160) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4192)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4192) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4224)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4224) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 4256)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4256) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4288)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4288) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4320)] = kernel[(((((((int)blockIdx.x) / 7) * 147456) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 138240)];
+ kernel_shared[(((int)threadIdx.x) + 4352)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4352) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 32) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4384)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4384) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 64) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4416)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4416) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 4448)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4448) / 144) * 4608)) + (rc_outer_outer * 144)) + ((((((int)threadIdx.x) + 128) % 144) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4480) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 16) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4512)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4512) / 144) * 4608)) + (rc_outer_outer * 144)) + ((int)threadIdx.x)) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 4544)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4544) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 80) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+ kernel_shared[(((int)threadIdx.x) + 4576)] = kernel[((((((((int)blockIdx.x) / 7) * 147456) + (((((int)threadIdx.x) + 4576) / 144) * 4608)) + (rc_outer_outer * 144)) + (((((int)threadIdx.x) + 112) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
__syncthreads();
- for (int rc_outer_inner = 0; rc_outer_inner < 2; ++rc_outer_inner) {
- for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
- for (int rx_outer_inner = 0; rx_outer_inner < 3; ++rx_outer_inner) {
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 144)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 288)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 1)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 2)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 3)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 4)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 5)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 6)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 432)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 9)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 153)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 297)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 81)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 82)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 83)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 84)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 85)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 86)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 87)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 441)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 18)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 162)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 306)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 162)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 163)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 164)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 165)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 166)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 167)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 168)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 450)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 27)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 171)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 315)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 243)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 244)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 245)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 246)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 247)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 248)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 249)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 459)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 36)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 180)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 324)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 324)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 325)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 326)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 327)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 328)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 329)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 330)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 468)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 45)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 189)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 333)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 405)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 406)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 407)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 408)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 409)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 410)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 411)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 477)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 54)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 198)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 342)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 486)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 487)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 488)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 489)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 490)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 491)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 492)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 486)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 63)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 207)]));
- conv2d_nchw[14] = (conv2d_nchw[14] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[15] = (conv2d_nchw[15] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[16] = (conv2d_nchw[16] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[17] = (conv2d_nchw[17] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[18] = (conv2d_nchw[18] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[19] = (conv2d_nchw[19] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[20] = (conv2d_nchw[20] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 351)]));
- conv2d_nchw[21] = (conv2d_nchw[21] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 567)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[22] = (conv2d_nchw[22] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 568)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[23] = (conv2d_nchw[23] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 569)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[24] = (conv2d_nchw[24] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 570)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[25] = (conv2d_nchw[25] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 571)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[26] = (conv2d_nchw[26] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 572)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- conv2d_nchw[27] = (conv2d_nchw[27] + (pad_temp_shared[(((((rc_outer_inner * 648) + (ry_outer_inner * 9)) + ((((int)threadIdx.x) % 7) * 9)) + rx_outer_inner) + 573)] * kernel_shared[((((((((int)threadIdx.x) / 7) * 576) + (rc_outer_inner * 72)) + (ry_outer_inner * 3)) + rx_outer_inner) + 495)]));
- }
- }
+ for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(ry_outer_inner * 9)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 1)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 27)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 28)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 54)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 55)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 81)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 82)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 108)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 109)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 135)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 136)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 162)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 163)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 189)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 190)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 216)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 217)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 243)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 244)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 270)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 271)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 297)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 298)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 324)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 325)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 351)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 352)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 378)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 379)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 405)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 406)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 1)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 28)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 55)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 82)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 109)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 136)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 163)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 190)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 217)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 244)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 271)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 298)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 325)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 352)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 379)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 406)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 2)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 29)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 56)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 83)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 110)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 137)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 164)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 191)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 218)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 245)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 272)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 299)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 326)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 353)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 380)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 407)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 3)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 30)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 57)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 84)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 111)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 138)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 165)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 192)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 219)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 246)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 273)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 300)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 327)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 354)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 381)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 408)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 4)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 31)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 58)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 85)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 112)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 139)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 166)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 193)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 220)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 247)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 274)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 301)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 328)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 355)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 382)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 409)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 5)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 7)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 32)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 34)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 59)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 61)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 86)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 88)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 113)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 115)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 140)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 142)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 167)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 169)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 194)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 196)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 221)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 223)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 248)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 250)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 275)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 277)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 302)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 304)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 329)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 331)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 356)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 358)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 383)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 385)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 410)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((ry_outer_inner * 9) + 412)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 6)] * kernel_shared[((((int)threadIdx.x) * 144) + (ry_outer_inner * 3))]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 7)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 1)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 8)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 2)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 33)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 9)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 34)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 10)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 35)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 11)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 60)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 18)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 61)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 19)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 62)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 20)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 87)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 27)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 88)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 28)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 89)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 29)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 114)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 36)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 115)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 37)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 116)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 38)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 141)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 45)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 142)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 46)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 143)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 47)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 168)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 54)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 169)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 55)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 170)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 56)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 195)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 63)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 196)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 64)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 197)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 65)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 222)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 72)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 223)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 73)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 224)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 74)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 249)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 81)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 250)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 82)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 251)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 83)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 276)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 90)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 277)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 91)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 278)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 92)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 303)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 99)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 304)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 100)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 305)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 101)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 330)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 108)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 331)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 109)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 332)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 110)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 357)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 117)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 358)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 118)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 359)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 119)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 384)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 126)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 385)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 127)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 386)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 128)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 411)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 135)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 412)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 136)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((ry_outer_inner * 9) + 413)] * kernel_shared[(((((int)threadIdx.x) * 144) + (ry_outer_inner * 3)) + 137)]));
}
}
- for (int i1_inner = 0; i1_inner < 4; ++i1_inner) {
- for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
- compute[(((((((int)blockIdx.x) * 3136) + ((((int)threadIdx.x) / 7) * 196)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[(((((int)blockIdx.x) * 64) + ((((int)threadIdx.x) / 7) * 4)) + i1_inner)]), 0.000000e+00f);
- }
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[(((((((int)blockIdx.x) / 7) * 1568) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[(((((int)blockIdx.x) / 7) * 32) + ((int)threadIdx.x))]), 0.000000e+00f);
}
}
</pre></div>
@@ -1512,7 +1987,7 @@ 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 38.539 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes 42.565 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 567213ac2d..110b21f8f1 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)
- 7.8691 7.8678 7.8797 7.8597 0.0082
+ 7.8563 7.8537 7.8646 7.8505 0.0060
</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 6.966 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 5.839 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 61cc08e0f7..8ca35a11c5 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)
- 761.8061 760.6130 765.3606 759.4448 2.5583
+ 758.6698 757.9524 760.1585 757.8985 1.0529
</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 40.846 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 39.693 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 c27d9a44c8..d6bcd76d6c 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -632,25 +632,26 @@ class Module:
@T.prim_func
def main(placeholder: T.Buffer((128, 256), "float32"), placeholder_1: T.Buffer((4916, 16, 1), "float32"), placeholder_2: T.Buffer((4916,), "int32"), placeholder_3: T.Buffer((33,), "int32"), placeholder_4: T.Buffer((128, 512), "float32"), compute: T.Buffer((128, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
- for i0_outer in T.parallel(8):
- compute_1 = T.allocate([256], "float32", "global")
- for i1_outer in range(32):
- compute_2 = T.Buffer((256,), data=compute_1)
- for i_inner_init, j_init in T.grid(16, 16):
- compute_2[i_inner_init * 16 + j_init] = T.float32(0)
- for elem_idx, i_inner, j in T.grid(placeholder_5[i1_outer + 1] - placeholder_5[i1_outer], 16, 16):
+ for i0_outer_i1_outer_fused in T.parallel(64):
+ compute_1 = T.allocate([1024], "float32", "global")
+ compute_2 = T.Buffer((1024,), data=compute_1)
+ for i_outer_inner, nb_j_inner in T.grid(16, 2):
+ for i_inner_init, j_init in T.grid(2, 16):
+ compute_2[i_outer_inner * 64 + i_inner_init * 32 + nb_j_inner * 16 + j_init] = T.float32(0)
+ for elem_idx, i_inner, j in T.grid(T.let(cse_var_1, i0_outer_i1_outer_fused % 16 * 2 + nb_j_inner, placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1]), 2, 16):
+ cse_var_1 = T.var("int32")
placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- if T.likely(elem_idx < placeholder_5[i1_outer + 1] - placeholder_5[i1_outer]):
- placeholder_6 = T.Buffer((78656,), data=placeholder_1.data)
- placeholder_7 = T.Buffer((32768,), data=placeholder.data)
- placeholder_8 = T.Buffer((4916,), "int32", data=placeholder_2.data)
- cse_var_1: T.int32 = i_inner * 16 + j
- compute_2[cse_var_1] = compute_2[cse_var_1] + placeholder_6[placeholder_5[i1_outer] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer * 4096 + i_inner * 256 + placeholder_8[placeholder_5[i1_outer] + elem_idx]], T.float32(0))
- for i0_inner in range(16):
- cse_var_2: T.int32 = i0_outer * 8192 + i0_inner * 512 + i1_outer * 16
- compute_3 = T.Buffer((65536,), data=compute.data)
- placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
- compute_3[cse_var_2:cse_var_2 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_2:cse_var_2 + 16], T.Broadcast(T.float32(0), 16))
+ cse_var_3: T.int32 = i0_outer_i1_outer_fused % 16 * 2 + nb_j_inner
+ cse_var_2: T.int32 = i_outer_inner * 64 + i_inner * 32 + nb_j_inner * 16 + j
+ placeholder_6 = T.Buffer((78656,), data=placeholder_1.data)
+ placeholder_7 = T.Buffer((32768,), data=placeholder.data)
+ placeholder_8 = T.Buffer((4916,), "int32", data=placeholder_2.data)
+ compute_2[cse_var_2] = compute_2[cse_var_2] + placeholder_6[placeholder_5[cse_var_3] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer_i1_outer_fused // 16 * 8192 + i_outer_inner * 512 + i_inner * 256 + placeholder_8[placeholder_5[cse_var_3] + elem_idx]], T.float32(0))
+ for i0_inner in range(32):
+ cse_var_4: T.int32 = i0_outer_i1_outer_fused // 16 * 16384 + i0_inner * 512 + i0_outer_i1_outer_fused % 16 * 32
+ compute_3 = T.Buffer((65536,), data=compute.data)
+ placeholder_5 = T.Buffer((65536,), data=placeholder_4.data)
+ compute_3[cse_var_4:cse_var_4 + 32] = T.max(compute_2[i0_inner * 32:i0_inner * 32 + 32] + placeholder_5[cse_var_4:cse_var_4 + 32], T.Broadcast(T.float32(0), 32))
</pre></div>
</div>
</div>
@@ -684,7 +685,7 @@ class Module:
<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: 1.522 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 2.117 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 ba5d94b09f..8ed446852f 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>00:29.987</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:43.161</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -349,11 +349,11 @@
</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>00:29.951</p></td>
+<td><p>00:43.129</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>
-<td><p>00:00.023</p></td>
+<td><p>00:00.019</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-cuda-py"><span class="std std-ref">Auto-tuning a Convolutional Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_cuda.py</span></code>)</p></td>
@@ -361,7 +361,7 @@
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tune_relay_arm.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-arm-py"><span class="std std-ref">Auto-tuning a Convolutional Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_arm.py</span></code>)</p></td>
-<td><p>00:00.004</p></td>
+<td><p>00:00.005</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="tune_relay_mobile_gpu.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-mobile-gpu-py"><span class="std std-ref">Auto-tuning a Convolutional Network for Mobile GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_mobile_gpu.py</span></code>)</p></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 2783454949..bd0ee276ab 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -690,9 +690,9 @@ 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, 8, 64]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2011889
-No: 2 GFLOPS: 124.09/124.09 result: MeasureResult(costs=(0.001865624372881356,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.5087034702301025, timestamp=1674771327.2719243) [('tile_f', [-1, 1, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 128, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1803160
-No: 3 GFLOPS: 0.00/124.09 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 4, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8495099
+No: 2 GFLOPS: 90.01/90.01 result: MeasureResult(costs=(0.002572052,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9165120124816895, timestamp=1674803118.8732517) [('tile_f', [-1, 8, 4, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,41072
+No: 3 GFLOPS: 0.00/90.01 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
@@ -814,10 +814,8 @@ 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, 32, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9198903
-No: 4 GFLOPS: 13.25/124.09 result: MeasureResult(costs=(0.017472539166666665,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.289174795150757, timestamp=1674771329.0655708) [('tile_f', [-1, 1, 16, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2979618
-No: 5 GFLOPS: 213.18/213.18 result: MeasureResult(costs=(0.001085961722972973,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.230830192565918, timestamp=1674771331.4487233) [('tile_f', [-1, 8, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9913017
-No: 6 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7487659
+No: 4 GFLOPS: 0.00/90.01 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
@@ -939,8 +937,8 @@ 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, 64, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4622248
-No: 7 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 64]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9392665
+No: 5 GFLOPS: 0.00/90.01 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
@@ -1062,8 +1060,8 @@ 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, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8189408
-No: 8 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 16, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 512]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2707796
+No: 6 GFLOPS: 0.00/90.01 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
@@ -1185,8 +1183,27 @@ 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, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 128, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9900518
-No: 9 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 4, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2806473
+No: 7 GFLOPS: 164.23/164.23 result: MeasureResult(costs=(0.0014096323918918918,), error_no=MeasureErrorNo.NO_ERROR, all_cost=10.716526985168457, timestamp=1674803132.9284441) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8844285
+No: 8 GFLOPS: 0.00/164.23 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, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9704202
+No: 9 GFLOPS: 0.00/164.23 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
@@ -1308,8 +1325,8 @@ 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, 1, 32, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8384330
-No: 10 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 32, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7838471
+No: 10 GFLOPS: 0.00/164.23 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
@@ -1431,9 +1448,8 @@ 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, 16, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 64]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,556949
-No: 11 GFLOPS: 4.01/213.18 result: MeasureResult(costs=(0.0577535055,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.6681346893310547, timestamp=1674771334.3050325) [('tile_f', [-1, 8, 2, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,288786
-No: 12 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 512, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,339294
+No: 11 GFLOPS: 0.00/164.23 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
@@ -1555,8 +1571,9 @@ 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, 4, 32, 2]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3924672
-No: 13 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 1, 256]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10117577
+No: 12 GFLOPS: 316.20/316.20 result: MeasureResult(costs=(0.0007321365147928994,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.7440016269683838, timestamp=1674803135.8509939) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4070442
+No: 13 GFLOPS: 0.00/316.20 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
@@ -1678,8 +1695,9 @@ 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, 4, 128, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7269291
-No: 14 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7471640
+No: 14 GFLOPS: 7.72/316.20 result: MeasureResult(costs=(0.02997332,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.854086637496948, timestamp=1674803141.9115317) [('tile_f', [-1, 4, 2, 8]), ('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, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2979825
+No: 15 GFLOPS: 0.00/316.20 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
@@ -1801,9 +1819,11 @@ 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, 8, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1991594
-No: 15 GFLOPS: 128.52/213.18 result: MeasureResult(costs=(0.001801294857142857,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2695696353912354, timestamp=1674771335.7931848) [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6320225
-No: 16 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10253255
+No: 16 GFLOPS: 111.35/316.20 result: MeasureResult(costs=(0.0020790924415584416,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.28901743888855, timestamp=1674803142.9225328) [('tile_f', [-1, 2, 4, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7074250
+No: 17 GFLOPS: 58.21/316.20 result: MeasureResult(costs=(0.003976914307692308,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2880961894989014, timestamp=1674803144.4809651) [('tile_f', [-1, 4, 16, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1668736
+No: 18 GFLOPS: 18.61/316.20 result: MeasureResult(costs=(0.012438851222222222,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.455913782119751, timestamp=1674803145.2864993) [('tile_f', [-1, 8, 1, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6210218
+No: 19 GFLOPS: 0.00/316.20 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
@@ -1925,8 +1945,8 @@ 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, 1, 8, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7339177
-No: 17 GFLOPS: 0.00/213.18 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 2, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9428211
+No: 20 GFLOPS: 0.00/316.20 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
@@ -2048,376 +2068,7 @@ 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, 8, 2, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2434446
-No: 18 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 128, 1, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9420727
-No: 19 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 128, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1482191
-No: 20 GFLOPS: 0.00/213.18 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:451
- 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:451
- 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, 2, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6155005
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 8, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8705647
</pre></div>
</div>
<p>Finally we can inspect the best config from log file, check correctness,
@@ -2456,9 +2107,9 @@ 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, 8, 16, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9913017
+[('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4070442
Finish loading 20 records
-Time cost of this operator: 0.001451
+Time cost of this operator: 0.001051
</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 3142c75ee5..b32cdd6524 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -647,10 +647,10 @@ 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 309.6 98.675 (1, 2, 10, 10, 3) 2 1 [309.6]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.115 0.993 (1, 6, 10, 10) 1 1 [3.115]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 1.041 0.332 (1, 1, 10, 10, 3) 1 1 [1.041]
-Total_time - 313.756 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 311.0 98.701 (1, 2, 10, 10, 3) 2 1 [311.0]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.116 0.989 (1, 6, 10, 10) 1 1 [3.116]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.977 0.31 (1, 1, 10, 10, 3) 1 1 [0.977]
+Total_time - 315.092 - - - - -
</pre></div>
</div>
</div>
@@ -702,10 +702,10 @@ 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 136.1 98.006 (1, 6, 10, 10, 1) 2 1 [136.1]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.788 1.288 (1, 6, 10, 10) 1 1 [1.788]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.981 0.706 (1, 1, 10, 10, 3) 1 1 [0.981]
-Total_time - 138.869 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 103.2 97.552 (1, 6, 10, 10, 1) 2 1 [103.2]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.75 1.654 (1, 6, 10, 10) 1 1 [1.75]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.84 0.794 (1, 3, 10, 10, 1) 1 1 [0.84]
+Total_time - 105.79 - - - - -
</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 31430f4ac7..e8771a72f7 100644
--- a/docs/how_to/work_with_microtvm/micro_pytorch.html
+++ b/docs/how_to/work_with_microtvm/micro_pytorch.html
@@ -454,8 +454,7 @@ 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]
- 61%|###### | 2.09M/3.42M [00:00<00:00, 17.7MB/s]
-100%|##########| 3.42M/3.42M [00:00<00:00, 27.6MB/s]
+100%|##########| 3.42M/3.42M [00:00<00:00, 60.7MB/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.
@@ -579,7 +578,7 @@ 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 11.795 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 9.814 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 6ac856f1b2..819e6f4524 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -524,7 +524,7 @@ 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/tmpl3ibk4mc/images/random'
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmp6krshqz2/images/random'
</pre></div>
</div>
</div>
@@ -584,8 +584,8 @@ 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], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [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/tmpl3ibk4mc/images/target contains 8144 images
-/tmp/tmpl3ibk4mc/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], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [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/tmp6krshqz2/images/target contains 8144 images
+/tmp/tmp6krshqz2/images/random contains 5000 images
</pre></div>
</div>
</div>
@@ -697,13 +697,13 @@ 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 - 48s - loss: 0.2199 - accuracy: 0.9236 - val_loss: 0.1252 - val_accuracy: 0.9588 - 48s/epoch - 146ms/step
+328/328 - 47s - loss: 0.2294 - accuracy: 0.9215 - val_loss: 0.1200 - val_accuracy: 0.9596 - 47s/epoch - 144ms/step
Epoch 2/3
-328/328 - 43s - loss: 0.0936 - accuracy: 0.9659 - val_loss: 0.1593 - val_accuracy: 0.9452 - 43s/epoch - 133ms/step
+328/328 - 44s - loss: 0.0962 - accuracy: 0.9643 - val_loss: 0.1094 - val_accuracy: 0.9626 - 44s/epoch - 133ms/step
Epoch 3/3
-328/328 - 43s - loss: 0.0722 - accuracy: 0.9730 - val_loss: 0.1136 - val_accuracy: 0.9611 - 43s/epoch - 132ms/step
+328/328 - 43s - loss: 0.0663 - accuracy: 0.9743 - val_loss: 0.2085 - val_accuracy: 0.9415 - 43s/epoch - 133ms/step
-<keras.callbacks.History object at 0x7f286aec7b90>
+<keras.callbacks.History object at 0x7fb3666e8e10>
</pre></div>
</div>
</div>
@@ -963,7 +963,7 @@ 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 38.806 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 22.236 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 2ac7885305..c8188906a9 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>06:56.419</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>06:36.598</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,23 +349,23 @@
</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>04:38.806</p></td>
+<td><p>04:22.236</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>01:11.795</p></td>
+<td><p>01:09.814</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>00:52.416</p></td>
+<td><p>00:51.727</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>00:09.523</p></td>
+<td><p>00:09.034</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>00:03.878</p></td>
+<td><p>00:03.787</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>
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 46776ba31c..916d6e2c83 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>00:46.429</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:45.047</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>00:33.678</p></td>
+<td><p>00:32.702</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>00:11.373</p></td>
+<td><p>00:10.617</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>00:01.372</p></td>
+<td><p>00:01.722</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 948715dc1f..b988e43109 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 0x7f285fef8f80>
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7fb3075403b0>
</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/sg_execution_times.html b/docs/how_to/work_with_schedules/sg_execution_times.html
index a071fc5b4a..e0be25cd18 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>00:07.531</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:06.262</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -349,23 +349,23 @@
</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>00:04.751</p></td>
+<td><p>00:03.654</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>00:01.378</p></td>
+<td><p>00:01.246</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>00:00.590</p></td>
+<td><p>00:00.576</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>00:00.586</p></td>
+<td><p>00:00.559</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>00:00.118</p></td>
+<td><p>00:00.119</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>
diff --git a/docs/how_to/work_with_schedules/tensorize.html b/docs/how_to/work_with_schedules/tensorize.html
index 294e7d780e..9dd31c9218 100644
--- a/docs/how_to/work_with_schedules/tensorize.html
+++ b/docs/how_to/work_with_schedules/tensorize.html
@@ -586,7 +586,7 @@ class Module:
def main(A: T.Buffer((1024, 64), "float32"), B: T.Buffer((512, 64), "float32"), C: T.Buffer((1024, 512), "float32")):
T.func_attr({"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True})
i = T.var("int32")
- T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpavgpahh8/input0.cc'\nsource_filename = \"/tmp/tmpavgpahh8/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 [...]
+ T.attr(T.iter_var(i, None, "DataPar", ""), "pragma_import_llvm", "; ModuleID = '/tmp/tmpro9643gx/input0.cc'\nsource_filename = \"/tmp/tmpro9643gx/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 [...]
for i, j_outer in T.grid(1024, 32):
T.call_extern("int32", "gemv_update", T.tvm_access_ptr(T.type_annotation("float32"), C.data, i * 512 + j_outer * 16, 16, 2), T.tvm_access_ptr(T.type_annotation("float32"), A.data, i * 64, 64, 1), T.tvm_access_ptr(T.type_annotation("float32"), B.data, j_outer * 1024, 1024, 1), 16, 64, 64)
</pre></div>
diff --git a/docs/reference/api/python/auto_scheduler.html b/docs/reference/api/python/auto_scheduler.html
index 7c08ce6085..d67bca2514 100644
--- a/docs/reference/api/python/auto_scheduler.html
+++ b/docs/reference/api/python/auto_scheduler.html
@@ -1615,7 +1615,7 @@ history states as starting point to perform Evolutionary Search).</p></li>
<dl class="py class">
<dt class="sig sig-object py" id="tvm.auto_scheduler.SketchPolicy">
-<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">SketchPolicy</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">program_cost_model</span></span><span class="o"><span class="pre">=</span></span><span class="defau [...]
+<em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">SketchPolicy</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">program_cost_model</span></span><span class="o"><span class="pre">=</span></span><span class="defau [...]
<dd><p>The search policy that searches in a hierarchical search space defined by sketches.
The policy randomly samples programs from the space defined by sketches and use evolutionary
search to fine-tune them.</p>
@@ -1899,7 +1899,7 @@ Candidates:
<dl class="py function">
<dt class="sig sig-object py" id="tvm.auto_scheduler.auto_schedule">
-<span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">auto_schedule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">search_policy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em clas [...]
+<span class="sig-prename descclassname"><span class="pre">tvm.auto_scheduler.</span></span><span class="sig-name descname"><span class="pre">auto_schedule</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">task</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">search_policy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em clas [...]
<dd><p>THIS API IS DEPRECATED.</p>
<p>Run auto scheduling search for a task.</p>
<dl class="field-list simple">
diff --git a/docs/reference/api/typedoc/classes/bytestreamreader.html b/docs/reference/api/typedoc/classes/bytestreamreader.html
index a15e33ffc3..308dc54933 100644
--- a/docs/reference/api/typedoc/classes/bytestreamreader.html
+++ b/docs/reference/api/typedoc/classes/bytestreamreader.html
@@ -119,7 +119,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -141,7 +141,7 @@
<div class="tsd-signature tsd-kind-icon">bytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Uint8Array</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L43">rpc_server.ts:43</a></li>
</ul>
</aside>
</section>
@@ -151,7 +151,7 @@
<div class="tsd-signature tsd-kind-icon">offset<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span><span class="tsd-signature-symbol"> = 0</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L42">rpc_server.ts:42</a></li>
</ul>
</aside>
</section>
@@ -168,7 +168,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L63">rpc_server.ts:63</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">Uint8Array</span></h4>
@@ -185,7 +185,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L49">rpc_server.ts:49</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -202,7 +202,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L57">rpc_server.ts:57</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
diff --git a/docs/reference/api/typedoc/classes/cachedcallstack.html b/docs/reference/api/typedoc/classes/cachedcallstack.html
index d0f97a2e31..5b198b68cd 100644
--- a/docs/reference/api/typedoc/classes/cachedcallstack.html
+++ b/docs/reference/api/typedoc/classes/cachedcallstack.html
@@ -144,7 +144,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L223">memory.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L223">memory.ts:223</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -172,7 +172,7 @@
<div class="tsd-signature tsd-kind-icon">temp<wbr>Args<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><a href="../interfaces/disposable.html" class="tsd-signature-type">Disposable</a><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = []</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L208">memory.ts:208</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L208">memory.ts:208</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -194,7 +194,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L312">memory.ts:312</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L312">memory.ts:312</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -226,7 +226,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L284">memory.ts:284</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L284">memory.ts:284</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -262,7 +262,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L388">memory.ts:388</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L388">memory.ts:388</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -300,7 +300,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L376">memory.ts:376</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L376">memory.ts:376</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -340,7 +340,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L267">memory.ts:267</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L267">memory.ts:267</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -373,7 +373,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L243">memory.ts:243</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L243">memory.ts:243</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -390,7 +390,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L321">memory.ts:321</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L321">memory.ts:321</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -422,7 +422,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L252">memory.ts:252</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L252">memory.ts:252</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -444,7 +444,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L359">memory.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L359">memory.ts:359</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -470,7 +470,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L342">memory.ts:342</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L342">memory.ts:342</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -496,7 +496,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L350">memory.ts:350</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L350">memory.ts:350</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -522,7 +522,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L326">memory.ts:326</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L326">memory.ts:326</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -548,7 +548,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L363">memory.ts:363</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L363">memory.ts:363</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -574,7 +574,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L346">memory.ts:346</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L346">memory.ts:346</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -600,7 +600,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L334">memory.ts:334</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L334">memory.ts:334</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
diff --git a/docs/reference/api/typedoc/classes/dldatatype.html b/docs/reference/api/typedoc/classes/dldatatype.html
index d628a543d7..99917893fd 100644
--- a/docs/reference/api/typedoc/classes/dldatatype.html
+++ b/docs/reference/api/typedoc/classes/dldatatype.html
@@ -119,7 +119,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L262">runtime.ts:262</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -147,7 +147,7 @@
<div class="tsd-signature tsd-kind-icon">bits<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L260">runtime.ts:260</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L260">runtime.ts:260</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">code<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L258">runtime.ts:258</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L258">runtime.ts:258</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -177,7 +177,7 @@
<div class="tsd-signature tsd-kind-icon">lanes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L262">runtime.ts:262</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L262">runtime.ts:262</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -199,7 +199,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L279">runtime.ts:279</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L279">runtime.ts:279</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -216,7 +216,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L270">runtime.ts:270</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L270">runtime.ts:270</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">string</span></h4>
diff --git a/docs/reference/api/typedoc/classes/dldevice.html b/docs/reference/api/typedoc/classes/dldevice.html
index 7ab39bcd32..4f6c6f3761 100644
--- a/docs/reference/api/typedoc/classes/dldevice.html
+++ b/docs/reference/api/typedoc/classes/dldevice.html
@@ -118,7 +118,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L202">runtime.ts:202</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L202">runtime.ts:202</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -146,7 +146,7 @@
<div class="tsd-signature tsd-kind-icon">device<wbr>Id<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L200">runtime.ts:200</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L200">runtime.ts:200</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -161,7 +161,7 @@
<div class="tsd-signature tsd-kind-icon">device<wbr>Type<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L198">runtime.ts:198</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L198">runtime.ts:198</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -183,7 +183,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L223">runtime.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L223">runtime.ts:223</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -205,7 +205,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L230">runtime.ts:230</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L230">runtime.ts:230</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">string</span></h4>
diff --git a/docs/reference/api/typedoc/classes/environment.html b/docs/reference/api/typedoc/classes/environment.html
index 7111cf512d..7095ec3134 100644
--- a/docs/reference/api/typedoc/classes/environment.html
+++ b/docs/reference/api/typedoc/classes/environment.html
@@ -125,7 +125,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L86">environment.ts:86</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L86">environment.ts:86</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -169,7 +169,7 @@
<aside class="tsd-sources">
<p>Implementation of <a href="../interfaces/libraryprovider.html">LibraryProvider</a>.<a href="../interfaces/libraryprovider.html#imports">imports</a></p>
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L70">environment.ts:70</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L70">environment.ts:70</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<div class="tsd-signature tsd-kind-icon">logger<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>msg<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">void</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L69">environment.ts:69</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L69">environment.ts:69</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -210,7 +210,7 @@
<div class="tsd-signature tsd-kind-icon">packedCFunc<wbr>Table<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">ctypes.FTVMWasmPackedCFunc</span><span class="tsd-signature-symbol"> | </span><span class="tsd-signature-type">undefined</span><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = [undefined,]</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L78">environment.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L78">environment.ts:78</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -228,7 +228,7 @@
<div class="tsd-signature tsd-kind-icon">packedCFunc<wbr>Table<wbr>Free<wbr>Id<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span><span class="tsd-signature-symbol"> = []</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L84">environment.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L84">environment.ts:84</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -250,7 +250,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/environment.ts#L105">environment.ts:105</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/environment.ts#L105">environment.ts:105</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/ffilibrary.html b/docs/reference/api/typedoc/classes/ffilibrary.html
index 42cabce7d6..8e584151d9 100644
--- a/docs/reference/api/typedoc/classes/ffilibrary.html
+++ b/docs/reference/api/typedoc/classes/ffilibrary.html
@@ -131,7 +131,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L49">runtime.ts:49</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L49">runtime.ts:49</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -156,7 +156,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L46">runtime.ts:46</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L46">runtime.ts:46</a></li>
</ul>
</aside>
</section>
@@ -166,7 +166,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L45">runtime.ts:45</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L45">runtime.ts:45</a></li>
</ul>
</aside>
</section>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">wasm32<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">boolean</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L44">runtime.ts:44</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L44">runtime.ts:44</a></li>
</ul>
</aside>
</section>
@@ -186,7 +186,7 @@
<div class="tsd-signature tsd-kind-icon">webGPUContext<span class="tsd-signature-symbol">:</span> <a href="webgpucontext.html" class="tsd-signature-type">WebGPUContext</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L47">runtime.ts:47</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L47">runtime.ts:47</a></li>
</ul>
</aside>
</section>
@@ -203,7 +203,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L76">runtime.ts:76</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L76">runtime.ts:76</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -226,7 +226,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L66">runtime.ts:66</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L66">runtime.ts:66</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -243,7 +243,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L84">runtime.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L84">runtime.ts:84</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <a href="cachedcallstack.html" class="tsd-signature-type">CachedCallStack</a></h4>
@@ -260,7 +260,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L95">runtime.ts:95</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L95">runtime.ts:95</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -283,7 +283,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L72">runtime.ts:72</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L72">runtime.ts:72</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
diff --git a/docs/reference/api/typedoc/classes/graphexecutor.html b/docs/reference/api/typedoc/classes/graphexecutor.html
index e1215f9204..72bc0348c1 100644
--- a/docs/reference/api/typedoc/classes/graphexecutor.html
+++ b/docs/reference/api/typedoc/classes/graphexecutor.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L583">runtime.ts:583</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L583">runtime.ts:583</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">module<span class="tsd-signature-symbol">:</span> <a href="module.html" class="tsd-signature-type">Module</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L579">runtime.ts:579</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L579">runtime.ts:579</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L654">runtime.ts:654</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L654">runtime.ts:654</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -224,7 +224,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L597">runtime.ts:597</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L597">runtime.ts:597</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -241,7 +241,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L631">runtime.ts:631</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L631">runtime.ts:631</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -279,7 +279,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L644">runtime.ts:644</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L644">runtime.ts:644</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -310,7 +310,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L621">runtime.ts:621</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L621">runtime.ts:621</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -332,7 +332,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L609">runtime.ts:609</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L609">runtime.ts:609</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/instance.html b/docs/reference/api/typedoc/classes/instance.html
index 14142bce64..36f0311231 100644
--- a/docs/reference/api/typedoc/classes/instance.html
+++ b/docs/reference/api/typedoc/classes/instance.html
@@ -139,7 +139,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L692">runtime.ts:692</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L692">runtime.ts:692</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -202,7 +202,7 @@
<div class="tsd-signature tsd-kind-icon">exports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">Function</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L684">runtime.ts:684</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L684">runtime.ts:684</a></li>
</ul>
</aside>
</section>
@@ -212,7 +212,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L683">runtime.ts:683</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L683">runtime.ts:683</a></li>
</ul>
</aside>
</section>
@@ -229,7 +229,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L932">runtime.ts:932</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L932">runtime.ts:932</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -260,7 +260,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L994">runtime.ts:994</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L994">runtime.ts:994</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -303,7 +303,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L924">runtime.ts:924</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L924">runtime.ts:924</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -341,7 +341,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L732">runtime.ts:732</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L732">runtime.ts:732</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -358,7 +358,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L952">runtime.ts:952</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L952">runtime.ts:952</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -402,7 +402,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L816">runtime.ts:816</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L816">runtime.ts:816</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -434,7 +434,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L1033">runtime.ts:1033</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -465,7 +465,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L846">runtime.ts:846</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L846">runtime.ts:846</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -497,7 +497,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L750">runtime.ts:750</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L750">runtime.ts:750</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -520,7 +520,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L1013">runtime.ts:1013</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -568,7 +568,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L789">runtime.ts:789</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L789">runtime.ts:789</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -608,7 +608,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L914">runtime.ts:914</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L914">runtime.ts:914</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -646,7 +646,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L1145">runtime.ts:1145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -698,7 +698,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L740">runtime.ts:740</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L740">runtime.ts:740</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -722,7 +722,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L868">runtime.ts:868</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L868">runtime.ts:868</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -754,7 +754,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L857">runtime.ts:857</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L857">runtime.ts:857</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -786,7 +786,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L940">runtime.ts:940</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L940">runtime.ts:940</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/memory.html b/docs/reference/api/typedoc/classes/memory.html
index cf7cce16e0..07eba0d859 100644
--- a/docs/reference/api/typedoc/classes/memory.html
+++ b/docs/reference/api/typedoc/classes/memory.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L40">memory.ts:40</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L40">memory.ts:40</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -152,7 +152,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Memory</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L32">memory.ts:32</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L32">memory.ts:32</a></li>
</ul>
</aside>
</section>
@@ -162,7 +162,7 @@
<div class="tsd-signature tsd-kind-icon">wasm32<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">boolean</span><span class="tsd-signature-symbol"> = true</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L33">memory.ts:33</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L33">memory.ts:33</a></li>
</ul>
</aside>
</section>
@@ -179,7 +179,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L154">memory.ts:154</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L154">memory.ts:154</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -210,7 +210,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L90">memory.ts:90</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L90">memory.ts:90</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -233,7 +233,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L97">memory.ts:97</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L97">memory.ts:97</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -256,7 +256,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L74">memory.ts:74</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L74">memory.ts:74</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -279,7 +279,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L81">memory.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L81">memory.ts:81</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -302,7 +302,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L104">memory.ts:104</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L104">memory.ts:104</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -325,7 +325,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L132">memory.ts:132</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L132">memory.ts:132</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -362,7 +362,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L145">memory.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L145">memory.ts:145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -393,7 +393,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L60">memory.ts:60</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L60">memory.ts:60</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -416,7 +416,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L67">memory.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L67">memory.ts:67</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -439,7 +439,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L53">memory.ts:53</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L53">memory.ts:53</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -462,7 +462,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L114">memory.ts:114</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L114">memory.ts:114</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -485,7 +485,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L124">memory.ts:124</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L124">memory.ts:124</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">number</span></h4>
@@ -502,7 +502,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/memory.ts#L175">memory.ts:175</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/memory.ts#L175">memory.ts:175</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/module.html b/docs/reference/api/typedoc/classes/module.html
index 2b8d74cbcb..04aa189795 100644
--- a/docs/reference/api/typedoc/classes/module.html
+++ b/docs/reference/api/typedoc/classes/module.html
@@ -124,7 +124,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L504">runtime.ts:504</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L504">runtime.ts:504</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -170,7 +170,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L502">runtime.ts:502</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L502">runtime.ts:502</a></li>
</ul>
</aside>
</section>
@@ -187,7 +187,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L516">runtime.ts:516</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L516">runtime.ts:516</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -204,7 +204,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L530">runtime.ts:530</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L530">runtime.ts:530</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -236,7 +236,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L561">runtime.ts:561</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L561">runtime.ts:561</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/ndarray.html b/docs/reference/api/typedoc/classes/ndarray.html
index 3d1d0e0b7d..5d3d8b7f0d 100644
--- a/docs/reference/api/typedoc/classes/ndarray.html
+++ b/docs/reference/api/typedoc/classes/ndarray.html
@@ -130,7 +130,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L304">runtime.ts:304</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L304">runtime.ts:304</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -158,7 +158,7 @@
<div class="tsd-signature tsd-kind-icon">device<span class="tsd-signature-symbol">:</span> <a href="dldevice.html" class="tsd-signature-type">DLDevice</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L297">runtime.ts:297</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L297">runtime.ts:297</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -173,7 +173,7 @@
<div class="tsd-signature tsd-kind-icon">dtype<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L293">runtime.ts:293</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L293">runtime.ts:293</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -188,7 +188,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L289">runtime.ts:289</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L289">runtime.ts:289</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -203,7 +203,7 @@
<div class="tsd-signature tsd-kind-icon">ndim<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L291">runtime.ts:291</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L291">runtime.ts:291</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -218,7 +218,7 @@
<div class="tsd-signature tsd-kind-icon">shape<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">Array</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">number</span><span class="tsd-signature-symbol">></span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L295">runtime.ts:295</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L295">runtime.ts:295</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -240,7 +240,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L370">runtime.ts:370</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L370">runtime.ts:370</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -273,7 +273,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L414">runtime.ts:414</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L414">runtime.ts:414</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -305,7 +305,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L355">runtime.ts:355</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L355">runtime.ts:355</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
@@ -322,7 +322,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L474">runtime.ts:474</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L474">runtime.ts:474</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -346,7 +346,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L443">runtime.ts:443</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L443">runtime.ts:443</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/packedfunccell.html b/docs/reference/api/typedoc/classes/packedfunccell.html
index deaa8a8516..c3488f7f70 100644
--- a/docs/reference/api/typedoc/classes/packedfunccell.html
+++ b/docs/reference/api/typedoc/classes/packedfunccell.html
@@ -122,7 +122,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L158">runtime.ts:158</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L158">runtime.ts:158</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -147,7 +147,7 @@
<div class="tsd-signature tsd-kind-icon">handle<span class="tsd-signature-symbol">:</span> <a href="../index.html#pointer" class="tsd-signature-type">Pointer</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L157">runtime.ts:157</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L157">runtime.ts:157</a></li>
</ul>
</aside>
</section>
@@ -164,7 +164,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L165">runtime.ts:165</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L165">runtime.ts:165</a></li>
</ul>
</aside>
<h4 class="tsd-returns-title">Returns <span class="tsd-signature-type">void</span></h4>
diff --git a/docs/reference/api/typedoc/classes/rpcserver.html b/docs/reference/api/typedoc/classes/rpcserver.html
index df731d011b..d62ad1c71f 100644
--- a/docs/reference/api/typedoc/classes/rpcserver.html
+++ b/docs/reference/api/typedoc/classes/rpcserver.html
@@ -115,7 +115,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L92">rpc_server.ts:92</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">get<wbr>Imports<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">Record</span><span class="tsd-signature-symbol"><</span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">, </span><span class="tsd-signature-type">unknown</span><span class="tsd-signat [...]
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L82">rpc_server.ts:82</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -201,7 +201,7 @@
<div class="tsd-signature tsd-kind-icon">key<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L78">rpc_server.ts:78</a></li>
</ul>
</aside>
</section>
@@ -211,7 +211,7 @@
<div class="tsd-signature tsd-kind-icon">logger<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol">(</span>msg<span class="tsd-signature-symbol">: </span><span class="tsd-signature-type">string</span><span class="tsd-signature-symbol">)</span><span class="tsd-signature-symbol"> => </span><span class="tsd-signature-type">void</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L81">rpc_server.ts:81</a></li>
</ul>
</aside>
<div class="tsd-type-declaration">
@@ -242,7 +242,7 @@
<div class="tsd-signature tsd-kind-icon">socket<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">WebSocket</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L79">rpc_server.ts:79</a></li>
</ul>
</aside>
</section>
@@ -252,7 +252,7 @@
<div class="tsd-signature tsd-kind-icon">state<span class="tsd-signature-symbol">:</span> <a href="../enums/rpcserverstate.html" class="tsd-signature-type">RPCServerState</a><span class="tsd-signature-symbol"> = RPCServerState.InitHeader</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L80">rpc_server.ts:80</a></li>
</ul>
</aside>
</section>
@@ -262,7 +262,7 @@
<div class="tsd-signature tsd-kind-icon">url<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L77">rpc_server.ts:77</a></li>
</ul>
</aside>
</section>
diff --git a/docs/reference/api/typedoc/classes/scalar.html b/docs/reference/api/typedoc/classes/scalar.html
index 9e8e0684d9..d700dffb93 100644
--- a/docs/reference/api/typedoc/classes/scalar.html
+++ b/docs/reference/api/typedoc/classes/scalar.html
@@ -112,7 +112,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L145">runtime.ts:145</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -137,7 +137,7 @@
<div class="tsd-signature tsd-kind-icon">dtype<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">string</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L145">runtime.ts:145</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L145">runtime.ts:145</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -152,7 +152,7 @@
<div class="tsd-signature tsd-kind-icon">value<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">number</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L143">runtime.ts:143</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L143">runtime.ts:143</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/classes/webgpucontext.html b/docs/reference/api/typedoc/classes/webgpucontext.html
index d27e4e9732..db43971ce8 100644
--- a/docs/reference/api/typedoc/classes/webgpucontext.html
+++ b/docs/reference/api/typedoc/classes/webgpucontext.html
@@ -120,7 +120,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L57">webgpu.ts:57</a></li>
</ul>
</aside>
<h4 class="tsd-parameters-title">Parameters</h4>
@@ -145,7 +145,7 @@
<div class="tsd-signature tsd-kind-icon">device<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-type">GPUDevice</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L50">webgpu.ts:50</a></li>
</ul>
</aside>
</section>
@@ -155,7 +155,7 @@
<div class="tsd-signature tsd-kind-icon">memory<span class="tsd-signature-symbol">:</span> <a href="memory.html" class="tsd-signature-type">Memory</a></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L51">webgpu.ts:51</a></li>
</ul>
</aside>
</section>
@@ -172,7 +172,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L84">webgpu.ts:84</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -209,7 +209,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L172">webgpu.ts:172</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L172">webgpu.ts:172</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
@@ -238,7 +238,7 @@
<li class="tsd-description">
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/webgpu.ts#L67">webgpu.ts:67</a></li>
</ul>
</aside>
<div class="tsd-comment tsd-typography">
diff --git a/docs/reference/api/typedoc/enums/argtypecode.html b/docs/reference/api/typedoc/enums/argtypecode.html
index 8cf0fd9a7b..b1512643bd 100644
--- a/docs/reference/api/typedoc/enums/argtypecode.html
+++ b/docs/reference/api/typedoc/enums/argtypecode.html
@@ -106,7 +106,7 @@
<div class="tsd-signature tsd-kind-icon">DLDevice<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 6</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L220">ctypes.ts:220</a></li>
</ul>
</aside>
</section>
@@ -116,7 +116,7 @@
<div class="tsd-signature tsd-kind-icon">Float<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 2</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L216">ctypes.ts:216</a></li>
</ul>
</aside>
</section>
@@ -126,7 +126,7 @@
<div class="tsd-signature tsd-kind-icon">Int<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 0</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L214">ctypes.ts:214</a></li>
</ul>
</aside>
</section>
@@ -136,7 +136,7 @@
<div class="tsd-signature tsd-kind-icon">Null<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 4</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L218">ctypes.ts:218</a></li>
</ul>
</aside>
</section>
@@ -146,7 +146,7 @@
<div class="tsd-signature tsd-kind-icon">TVMBytes<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 12</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L226">ctypes.ts:226</a></li>
</ul>
</aside>
</section>
@@ -156,7 +156,7 @@
<div class="tsd-signature tsd-kind-icon">TVMDLTensor<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 7</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L221">ctypes.ts:221</a></li>
</ul>
</aside>
</section>
@@ -166,7 +166,7 @@
<div class="tsd-signature tsd-kind-icon">TVMData<wbr>Type<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 5</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L219">ctypes.ts:219</a></li>
</ul>
</aside>
</section>
@@ -176,7 +176,7 @@
<div class="tsd-signature tsd-kind-icon">TVMModule<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 9</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L223">ctypes.ts:223</a></li>
</ul>
</aside>
</section>
@@ -186,7 +186,7 @@
<div class="tsd-signature tsd-kind-icon">TVMNDArray<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 13</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L227">ctypes.ts:227</a></li>
</ul>
</aside>
</section>
@@ -196,7 +196,7 @@
<div class="tsd-signature tsd-kind-icon">TVMObject<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 8</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L222">ctypes.ts:222</a></li>
</ul>
</aside>
</section>
@@ -206,7 +206,7 @@
<div class="tsd-signature tsd-kind-icon">TVMObjectRValue<wbr>Ref<wbr>Arg<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 14</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L228">ctypes.ts:228</a></li>
</ul>
</aside>
</section>
@@ -216,7 +216,7 @@
<div class="tsd-signature tsd-kind-icon">TVMOpaque<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 3</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L217">ctypes.ts:217</a></li>
</ul>
</aside>
</section>
@@ -226,7 +226,7 @@
<div class="tsd-signature tsd-kind-icon">TVMPacked<wbr>Func<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 10</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L224">ctypes.ts:224</a></li>
</ul>
</aside>
</section>
@@ -236,7 +236,7 @@
<div class="tsd-signature tsd-kind-icon">TVMStr<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 11</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L225">ctypes.ts:225</a></li>
</ul>
</aside>
</section>
@@ -246,7 +246,7 @@
<div class="tsd-signature tsd-kind-icon">UInt<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 1</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/ctypes.ts#L215">ctypes.ts:215</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/ctypes.ts#L215">ctypes.ts:215</a></li>
</ul>
</aside>
</section>
diff --git a/docs/reference/api/typedoc/enums/aynccallbackcode.html b/docs/reference/api/typedoc/enums/aynccallbackcode.html
index c4ecb08aeb..dff5861a33 100644
--- a/docs/reference/api/typedoc/enums/aynccallbackcode.html
+++ b/docs/reference/api/typedoc/enums/aynccallbackcode.html
@@ -93,7 +93,7 @@
<div class="tsd-signature tsd-kind-icon">k<wbr>Exception<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 5</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L676">runtime.ts:676</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L676">runtime.ts:676</a></li>
</ul>
</aside>
</section>
@@ -103,7 +103,7 @@
<div class="tsd-signature tsd-kind-icon">k<wbr>Return<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 4</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L675">runtime.ts:675</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L675">runtime.ts:675</a></li>
</ul>
</aside>
</section>
diff --git a/docs/reference/api/typedoc/enums/dldatatypecode.html b/docs/reference/api/typedoc/enums/dldatatypecode.html
index c40c7aedbb..d30669609b 100644
--- a/docs/reference/api/typedoc/enums/dldatatypecode.html
+++ b/docs/reference/api/typedoc/enums/dldatatypecode.html
@@ -95,7 +95,7 @@
<div class="tsd-signature tsd-kind-icon">Float<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 2</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L242">runtime.ts:242</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L242">runtime.ts:242</a></li>
</ul>
</aside>
</section>
@@ -105,7 +105,7 @@
<div class="tsd-signature tsd-kind-icon">Int<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 0</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L240">runtime.ts:240</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L240">runtime.ts:240</a></li>
</ul>
</aside>
</section>
@@ -115,7 +115,7 @@
<div class="tsd-signature tsd-kind-icon">Opaque<wbr>Handle<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 3</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L243">runtime.ts:243</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L243">runtime.ts:243</a></li>
</ul>
</aside>
</section>
@@ -125,7 +125,7 @@
<div class="tsd-signature tsd-kind-icon">UInt<span class="tsd-signature-symbol">:</span> <span class="tsd-signature-symbol"> = 1</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/runtime.ts#L241">runtime.ts:241</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/runtime.ts#L241">runtime.ts:241</a></li>
</ul>
</aside>
</section>
diff --git a/docs/reference/api/typedoc/enums/rpcserverstate.html b/docs/reference/api/typedoc/enums/rpcserverstate.html
index 0f0509b436..0b96e44627 100644
--- a/docs/reference/api/typedoc/enums/rpcserverstate.html
+++ b/docs/reference/api/typedoc/enums/rpcserverstate.html
@@ -90,7 +90,7 @@
<div class="tsd-signature tsd-kind-icon">Init<wbr>Header<span class="tsd-signature-symbol">:</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L27">rpc_server.ts:27</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L27">rpc_server.ts:27</a></li>
</ul>
</aside>
</section>
@@ -100,7 +100,7 @@
<div class="tsd-signature tsd-kind-icon">Init<wbr>Header<wbr>Key<span class="tsd-signature-symbol">:</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L28">rpc_server.ts:28</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L28">rpc_server.ts:28</a></li>
</ul>
</aside>
</section>
@@ -110,7 +110,7 @@
<div class="tsd-signature tsd-kind-icon">Init<wbr>Server<span class="tsd-signature-symbol">:</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L29">rpc_server.ts:29</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L29">rpc_server.ts:29</a></li>
</ul>
</aside>
</section>
@@ -120,7 +120,7 @@
<div class="tsd-signature tsd-kind-icon">Receive<wbr>Packet<wbr>Body<span class="tsd-signature-symbol">:</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L32">rpc_server.ts:32</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L32">rpc_server.ts:32</a></li>
</ul>
</aside>
</section>
@@ -130,7 +130,7 @@
<div class="tsd-signature tsd-kind-icon">Receive<wbr>Packet<wbr>Header<span class="tsd-signature-symbol">:</span></div>
<aside class="tsd-sources">
<ul>
- <li>Defined in <a href="https://github.com/apache/tvm/blob/4ddb87631/web/src/rpc_server.ts#L31">rpc_server.ts:31</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/18b7dc1dd/web/src/rpc_server.ts#L31">rpc_server.ts:31</a></li>
</ul>
</aside>
</section>
@@ -140,7 +140,7 @@
<div class="tsd-signature tsd-kind-icon">Wait<wbr>For<wbr>Callback<span class="tsd-signature-symbol">:</span></div>
... 1446 lines suppressed ...