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 2023/05/02 05:19:57 UTC
[tvm-site] branch asf-site updated: deploying docs (apache/tvm@c1e88b365a240a99452b33f0fe30776e01629125)
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 2f6ffde1d4 deploying docs (apache/tvm@c1e88b365a240a99452b33f0fe30776e01629125)
2f6ffde1d4 is described below
commit 2f6ffde1d4f1b3789223e7ab85642c38d9aff75c
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Tue May 2 05:19:50 2023 +0000
deploying docs (apache/tvm@c1e88b365a240a99452b33f0fe30776e01629125)
---
docs/_images/sphx_glr_micro_train_001.png | Bin 346533 -> 341097 bytes
docs/_images/sphx_glr_micro_train_thumb.png | Bin 24178 -> 24317 bytes
.../how_to/compile_models/from_darknet.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_paddle.rst.txt | 5 -
.../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_model_on_adreno_tvmc.rst.txt | 2 +-
.../deploy_models/deploy_model_on_android.rst.txt | 2 +-
.../deploy_object_detection_pytorch.rst.txt | 4 +-
.../deploy_models/deploy_prequantized.rst.txt | 6 +-
.../deploy_prequantized_tflite.rst.txt | 4 +-
.../how_to/deploy_models/deploy_quantized.rst.txt | 2 +-
.../deploy_models/deploy_ssd_gluoncv.rst.txt | 4 +-
.../deploy_models/sg_execution_times.rst.txt | 22 +-
.../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 | 2078 ++++++++------------
.../tune_network_cuda.rst.txt | 4 +-
.../tune_network_x86.rst.txt | 4 +-
.../tune_sparse_x86.rst.txt | 38 +-
.../tune_with_autotvm/sg_execution_times.rst.txt | 4 +-
.../tune_with_autotvm/tune_conv2d_cuda.rst.txt | 843 +++++++-
.../work_with_microtvm/micro_autotune.rst.txt | 18 +-
.../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 | 14 +-
.../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 | 16 +-
.../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 | 4 +-
docs/_sources/tutorial/autotvm_matmul_x86.rst.txt | 20 +-
docs/_sources/tutorial/autotvm_relay_x86.rst.txt | 60 +-
.../tutorial/cross_compilation_and_rpc.rst.txt | 2 +-
docs/_sources/tutorial/intro_topi.rst.txt | 2 +-
docs/_sources/tutorial/sg_execution_times.rst.txt | 18 +-
.../tutorial/tensor_expr_get_started.rst.txt | 46 +-
docs/commit_hash | 2 +-
docs/how_to/compile_models/from_darknet.html | 2 +-
docs/how_to/compile_models/from_mxnet.html | 2 +-
docs/how_to/compile_models/from_oneflow.html | 15 +-
docs/how_to/compile_models/from_paddle.html | 1 -
docs/how_to/compile_models/from_pytorch.html | 18 +-
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_adreno_tvmc.html | 27 +-
.../deploy_models/deploy_model_on_android.html | 2 +-
.../deploy_object_detection_pytorch.html | 65 +-
docs/how_to/deploy_models/deploy_prequantized.html | 9 +-
.../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 | 40 +-
docs/how_to/deploy_models/sg_execution_times.html | 22 +-
.../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 | 2078 ++++++++------------
.../tune_with_autoscheduler/tune_network_cuda.html | 4 +-
.../tune_with_autoscheduler/tune_network_x86.html | 4 +-
.../tune_with_autoscheduler/tune_sparse_x86.html | 38 +-
.../tune_with_autotvm/sg_execution_times.html | 4 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.html | 843 +++++++-
docs/how_to/work_with_microtvm/micro_autotune.html | 18 +-
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 | 14 +-
.../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 | 16 +-
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 +-
docs/reference/api/typedoc/classes/instance.html | 58 +-
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 +-
.../api/typedoc/classes/runtimecontext.html | 22 +-
docs/reference/api/typedoc/classes/scalar.html | 6 +-
docs/reference/api/typedoc/classes/tvmarray.html | 16 +-
docs/reference/api/typedoc/classes/tvmobject.html | 12 +-
.../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 | 124 +-
.../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 | 4 +-
docs/tutorial/autotvm_matmul_x86.html | 20 +-
docs/tutorial/autotvm_relay_x86.html | 270 +--
docs/tutorial/cross_compilation_and_rpc.html | 2 +-
docs/tutorial/intro_topi.html | 2 +-
docs/tutorial/sg_execution_times.html | 18 +-
docs/tutorial/tensor_expr_get_started.html | 46 +-
131 files changed, 4044 insertions(+), 3613 deletions(-)
diff --git a/docs/_images/sphx_glr_micro_train_001.png b/docs/_images/sphx_glr_micro_train_001.png
index 1612aabc2c..60e9d514ac 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 b25d3cb0c8..537d31c680 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 1be5eb3783..1247a5ac5a 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 27.616 seconds)
+ **Total running time of the script:** ( 1 minutes 24.849 seconds)
.. _sphx_glr_download_how_to_compile_models_from_darknet.py:
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 41dafdf0c6..01c9c1815f 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.zip995aba3a-7257-4a91-8e80-24528214d098 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+ Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipf60f7196-0da9-4273-afb2-2b6a6e4c8534 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 b9b9a861e3..8975dc598e 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]
19%|#9 | 8.00M/41.5M [00:00<00:01, 31.5MB/s]
39%|###8 | 16.0M/41.5M [00:00<00:00, 35.7MB/s]
54%|#####3 | 22.3M/41.5M [00:00<00:00, 36.6MB/s]
62%|######2 | 25.9M/41.5M [00:00<00:00, 32.8MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 37.5MB/s]
90%|########9 | 37.2M/41.5M [00:01<00:00, 41.2MB/s]
100%|#########9| 41.3M/41.5M [00:01<00:00, 40.1MB/s]
100%|##########| 41.5M/41.5M [00:01<00:00, 37.7MB/s]
+
0%| | 0.00/41.5M [00:00<?, ?B/s]
15%|#5 | 6.33M/41.5M [00:00<00:00, 60.2MB/s]
29%|##9 | 12.1M/41.5M [00:00<00:00, 53.0MB/s]
41%|####1 | 17.2M/41.5M [00:00<00:00, 32.9MB/s]
58%|#####7 | 24.0M/41.5M [00:00<00:00, 39.3MB/s]
77%|#######7 | 32.0M/41.5M [00:00<00:00, 47.2MB/s]
92%|#########2| 38.3M/41.5M [00:00<00:00, 48.6MB/s]
100%|##########| 41.5M/41.5M [00:00<00:00, 43.9MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_paddle.rst.txt b/docs/_sources/how_to/compile_models/from_paddle.rst.txt
index ee9e7ab431..8df75c9e33 100644
--- a/docs/_sources/how_to/compile_models/from_paddle.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_paddle.rst.txt
@@ -207,11 +207,6 @@ Look up prediction top 1 index in 1000 class synset.
-.. rst-class:: sphx-glr-timing
-
- **Total running time of the script:** ( 1 minutes 1.022 seconds)
-
-
.. _sphx_glr_download_how_to_compile_models_from_paddle.py:
.. only:: html
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 7e6dd41cd8..9813d5c1a1 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]
14%|#4 | 6.30M/44.7M [00:00<00:00, 48.1MB/s]
24%|##4 | 10.9M/44.7M [00:00<00:00, 39.6MB/s]
36%|###5 | 16.0M/44.7M [00:00<00:00, 32.4MB/s]
54%|#####3 | 24.0M/44.7M [00:00<00:00, 41.9MB/s]
63%|######3 | 28.2M/44.7M [00:00<00:00, 42.1MB/s]
72%|#######2 | 32.3M/44.7M [00:01<00:00, 22.0MB/s]
86%|########5 | 38.3M/44.7M [00:01<00:00, 26.1MB/s]
97%|#########6| 43.1M/44.7M [00:01<00:00, 30.5MB/s]
100%|##########| 44.7M/44.7M [00:01<00:00, 30.4MB/s]
+
0%| | 0.00/44.7M [00:00<?, ?B/s]
14%|#4 | 6.30M/44.7M [00:00<00:00, 42.1MB/s]
23%|##3 | 10.3M/44.7M [00:00<00:01, 32.5MB/s]
32%|###2 | 14.3M/44.7M [00:00<00:01, 29.8MB/s]
38%|###8 | 17.2M/44.7M [00:00<00:00, 29.3MB/s]
54%|#####3 | 24.0M/44.7M [00:00<00:00, 38.6MB/s]
68%|######7 | 30.3M/44.7M [00:00<00:00, 35.6MB/s]
76%|#######5 | 33.8M/44.7M [00:01<00:00, 33.8MB/s]
90%|########9 | 40.0M/44.7M [00:01<00:00, 36.9MB/s]
100%|##########| 44.7M/44.7M [00:01<00:00, 38.4MB/s]
diff --git a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
index 2490279f6d..1d9cad1d8c 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -430,7 +430,7 @@ Run the corresponding model on tensorflow
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 36.123 seconds)
+ **Total running time of the script:** ( 1 minutes 33.495 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 81e117294a..edabe43f84 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
=================
-**07:18.513** total execution time for **how_to_compile_models** files:
+**07:06.376** total execution time for **how_to_compile_models** files:
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:36.123 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:33.495 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:27.616 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``) | 01:24.849 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 01:01.022 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``) | 00:58.912 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:40.162 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``) | 00:39.303 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:34.765 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``) | 00:33.564 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:31.741 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``) | 00:31.068 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:30.066 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``) | 00:29.804 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:28.300 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``) | 00:27.221 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:25.827 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``) | 00:25.335 | 0.0 MB |
+-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.891 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``) | 00:02.824 | 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 bea69deb4f..f3130c34ee 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
@@ -673,18 +673,13 @@ 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)
- 2855.3580 2855.7379 2858.1845 2852.2353 1.7204
+ 2539.4086 2539.2871 2541.9247 2538.0618 1.0440
-.. rst-class:: sphx-glr-timing
-
- **Total running time of the script:** ( 1 minutes 0.147 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_adreno_tvmc.rst.txt b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno_tvmc.rst.txt
index da950b9041..c7aafb9ff6 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno_tvmc.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno_tvmc.rst.txt
@@ -127,7 +127,7 @@ Make a Keras Resnet50 Model
.. code-block:: none
Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels.h5
-
8192/102967424 [..............................] - ETA: 0s
5906432/102967424 [>.............................] - ETA: 0s
8380416/102967424 [=>............................] - ETA: 1s
15024128/102967424 [===>..........................] - ETA: 1s
16769024/102967424 [===>..........................] - ETA: 1s
23412736/102967424 [=====>........................] - ETA: 1s
25157632/102967424 [======>.......................] - ETA: 2s
33546240/102967424 [========>.....................] - ETA: 1s
40189952/102967424 [==========>...................] - ETA: 2s
41934848/102967424 [===========>..................] - ETA: 2s
50323456/102967424 [=============>................] - ETA: 1s
53395456/102967424 [==============>...............] - ETA: 1s
56967168/102967424 [===============>..............] - ETA: 1s
58712064/102967424 [================>.............] - ETA: 1s
65355776/102967424 [==================>...........] - ETA: 1s
67100672/102967424 [==================>...........] -
ETA: 1s
69296128/102967424 [===================>..........] - ETA: 1s
73744384/102967424 [====================>.........] - ETA: 0s
75489280/102967424 [====================>.........] - ETA: 0s
82124800/102967424 [======================>.......] - ETA: 0s
83877888/102967424 [=======================>......] - ETA: 0s
90521600/102967424 [=========================>....] - ETA: 0s
92266496/102967424 [=========================>....] - ETA: 0s
98910208/102967424
[===========================>..] - ETA: 0s
100646912/102967424 [============================>.] - ETA: 0s
102850560/102967424 [============================>.] - ETA: 0s
102967424/102967424 [==============================] - 3s 0us/step
+
8192/102967424 [..............................] - ETA: 0s
131072/102967424 [..............................] - ETA: 1:12
6635520/102967424 [>.............................] - ETA: 2s
8380416/102967424 [=>............................] - ETA: 3s
15024128/102967424 [===>..........................] - ETA: 1s
16769024/102967424 [===>..........................] - ETA: 2s
16908288/102967424 [===>..........................] - ETA: 2s
23412736/102967424 [=====>........................] - ETA: 2s
25157632/102967424 [======>.......................] - ETA: 2s
31203328/102967424 [========>.....................] - ETA: 1s
33546240/102967424 [========>.....................] - ETA: 1s
40026112/102967424 [==========>...................] - ETA: 1s
41934848/102967424 [===========>..................] - ETA: 1s
48578560/102967424 [=============>................] - ETA: 1s
50323456/102967424 [=============>................] - ETA: 1s
56967168/102967424 [===============>...........
...] - ETA: 1s
58712064/102967424 [================>.............] - ETA: 1s
65912832/102967424 [==================>...........] - ETA: 0s
67100672/102967424 [==================>...........] - ETA: 0s
69296128/102967424 [===================>..........] - ETA: 0s
74522624/102967424 [====================>.........] - ETA: 0s
75489280/102967424 [====================>.........] - ETA: 0s
82124800/102967424 [======================>.......] - ETA: 0s
83877888/102
967424 [=======================>......] - ETA: 0s
90521600/102967424 [=========================>....] - ETA: 0s
92266496/102967424 [=========================>....] - ETA: 0s
98910208/102967424 [===========================>..] - ETA: 0s
100794368/102967424 [============================>.] - ETA: 0s
102850560/102967424 [============================>.] - ETA: 0s
102967424/102967424 [==============================] - 3s 0us/step
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 ea08bc6fec..a90822973c 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)
- 16.8965 17.1268 17.2892 16.0530 0.4442
+ 15.8141 15.8113 15.9127 15.7327 0.0569
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 ba7e2364e5..d3c7342d49 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]
4%|3 | 6.30M/170M [00:00<00:03, 56.2MB/s]
7%|6 | 11.7M/170M [00:00<00:03, 46.6MB/s]
10%|9 | 16.2M/170M [00:00<00:05, 29.1MB/s]
13%|#3 | 22.3M/170M [00:00<00:04, 37.4MB/s]
16%|#5 | 26.5M/170M [00:00<00:04, 32.2MB/s]
19%|#8 | 32.0M/170M [00:01<00:05, 27.6MB/s]
24%|##3 | 40.0M/170M [00:01<00:03, 36.4MB/s]
27%|##7 | 46.3M/170M [00:01<00:03, 39.3MB/s]
30%|##9 | 50.5M/170M [00:01<00:03, 32.7MB/s]
33%|###2 | 56.0M/170M [00:01<00:03, 34.8MB/s]
38%|###7 | 64.0M/170M [00:01<00:02, 38.8MB/s]
42%|####2 | 72.0M/170M [00:02<00:02, 43.8MB/s]
46%|####6 | 78.3M/170M [00:02<00:02, 40.3MB/s]
48%|####8 | 82.3M/170M [00:02<00:02, 37.0MB/s]
51%|##### | 86.3M/170M [00:02<00:02, 33.4MB/s]
53%|#####2 | 89.6M/170M [00:02<00:02, 32.8MB/s]
56%|#####5 | 94.3M/170M [00:02<00:02, 34.1MB/
s]
57%|#####7 | 97.6M/170M [00:02<00:02, 33.2MB/s]
60%|###### | 102M/170M [00:03<00:02, 32.4MB/s]
62%|######2 | 105M/170M [00:03<00:02, 30.1MB/s]
66%|######5 | 112M/170M [00:03<00:01, 34.3MB/s]
71%|####### | 120M/170M [00:03<00:01, 43.3MB/s]
75%|#######5 | 128M/170M [00:03<00:01, 41.8MB/s]
80%|######## | 136M/170M [00:03<00:00, 38.2MB/s]
85%|########4 | 144M/170M [00:04<00:00, 40.0MB/s]
88%|########8 | 150M/170M [00:04<00:00, 43.5MB/s]
91%|#########1| 155M/170M [00:04<00:00, 35.2MB/s]
93%|#########3| 158M/170M [00:04<00:00, 30.9MB/s]
95%|#########5| 162M/170M [00:04<00:00, 28.5MB/s]
98%|#########7| 166M/170M [00:04<00:00, 30.6MB/s]
100%|##########| 170M/170M [00:04<00:00, 35.6MB/s]
+
0%| | 0.00/170M [00:00<?, ?B/s]
4%|3 | 6.30M/170M [00:00<00:07, 23.3MB/s]
5%|5 | 8.53M/170M [00:00<00:07, 22.9MB/s]
8%|8 | 14.3M/170M [00:00<00:09, 17.4MB/s]
11%|# | 18.5M/170M [00:00<00:07, 22.3MB/s]
14%|#4 | 24.0M/170M [00:01<00:05, 26.0MB/s]
19%|#8 | 32.0M/170M [00:01<00:04, 31.4MB/s]
23%|##2 | 38.6M/170M [00:01<00:03, 38.6MB/s]
25%|##5 | 42.9M/170M [00:01<00:04, 31.9MB/s]
28%|##8 | 48.0M/170M [00:01<00:03, 32.1MB/s]
33%|###2 | 56.0M/170M [00:01<00:03, 35.4MB/s]
38%|###7 | 64.0M/170M [00:02<00:02, 40.3MB/s]
42%|####2 | 72.0M/170M [00:02<00:02, 45.3MB/s]
46%|####6 | 78.3M/170M [00:02<00:02, 41.9MB/s]
49%|####8 | 82.5M/170M [00:02<00:02, 36.8MB/s]
52%|#####1 | 88.0M/170M [00:02<00:02, 34.5MB/s]
56%|#####5 | 94.3M/170M [00:02<00:02, 36.6MB/s]
58%|#####7 | 97.9M/170M [00:03<00:02, 35.6MB/
s]
60%|###### | 102M/170M [00:03<00:02, 35.0MB/s]
62%|######2 | 106M/170M [00:03<00:02, 33.6MB/s]
66%|######5 | 112M/170M [00:03<00:01, 35.1MB/s]
71%|####### | 120M/170M [00:03<00:01, 38.2MB/s]
74%|#######4 | 126M/170M [00:03<00:01, 42.7MB/s]
77%|#######6 | 130M/170M [00:03<00:00, 42.3MB/s]
80%|######## | 136M/170M [00:04<00:00, 38.8MB/s]
85%|########4 | 144M/170M [00:04<00:00, 42.1MB/s]
88%|########8 | 150M/170M [00:04<00:00, 42.9MB/s]
91%|######### | 154M/170M [00:04<00:00, 40.1MB/s]
93%|#########3| 158M/170M [00:04<00:00, 39.7MB/s]
95%|#########5| 162M/170M [00:04<00:00, 33.7MB/s]
98%|#########7| 166M/170M [00:05<00:00, 31.1MB/s]
100%|#########9| 169M/170M [00:05<00:00, 30.9MB/s]
100%|##########| 170M/170M [00:05<00:00, 34.8MB/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 52.109 seconds)
+ **Total running time of the script:** ( 3 minutes 41.966 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 6a0db86640..ffebbab914 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, 37.1MB/s]
90%|########9 | 12.2M/13.6M [00:00<00:00, 38.0MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 41.4MB/s]
+
0%| | 0.00/13.6M [00:00<?, ?B/s]
59%|#####8 | 7.99M/13.6M [00:00<00:00, 73.5MB/s]
100%|##########| 13.6M/13.6M [00:00<00:00, 107MB/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)
- 90.6979 90.5879 95.3150 90.3424 0.5525
+ 90.6953 90.6578 91.6423 90.3184 0.2578
@@ -458,7 +458,7 @@ TODO
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 21.891 seconds)
+ **Total running time of the script:** ( 1 minutes 18.908 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 e060a73b8f..497083d2fc 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)
- 120.5770 120.5138 122.3097 119.4893 0.4387
+ 120.9231 120.8937 121.9398 120.0892 0.3639
@@ -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 32.983 seconds)
+ **Total running time of the script:** ( 2 minutes 40.736 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 14461c6426..0d0e2926a5 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 56.443 seconds)
+ **Total running time of the script:** ( 1 minutes 52.802 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 bfa7e3a73d..04a971997f 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 | 3891/132723 [00:00<00:03, 38706.09KB/s]
9%|8 | 11899/132723 [00:00<00:01, 62987.25KB/s]
14%|#3 | 18203/132723 [00:00<00:02, 46444.99KB/s]
20%|## | 26621/132723 [00:00<00:01, 58685.28KB/s]
25%|##4 | 32995/132723 [00:00<00:02, 43196.03KB/s]
31%|###1 | 41420/132723 [00:00<00:01, 53244.53KB/s]
37%|###7 | 49140/132723 [00:01<00:01, 45672.94KB/s]
43%|####3 | 57536/132723 [00:01<00:01, 54148.48KB/s]
49%|####9 | 65522/132723 [00:01<00:01, 46448.10KB/s]
56%|#####5 | 73882/132723 [00:01<00:01, 54213.92KB/s]
61%|###### | 80692/132723 [00:01<00:01, 47423.74KB/s]
65%|######4 | 86191/132723 [00:01<00:01, 42013.67KB/s]
69%|######8 | 90955/132723 [00:01<00:00, 42956.29KB/s]
74%|#######3 | 97594/132723 [00:02<00:00, 42894.85KB/s]
77%|#######6 | 102186/132723 [00:02<00:00, 34811.40KB/s]
83%|########3
| 110653/132723 [00:02<00:00, 44960.92KB/s]
87%|########7 | 115900/132723 [00:02<00:00, 35105.08KB/s]
94%|#########3| 124342/132723 [00:02<00:00, 44718.36KB/s]
100%|##########| 132723/132723 [00:02<00:00, 53157.78KB/s]
100%|##########| 132723/132723 [00:02<00:00, 46862.48KB/s]
+
0%| | 0/132723 [00:00<?, ?KB/s]
5%|5 | 7222/132723 [00:00<00:01, 72208.53KB/s]
12%|#1 | 15768/132723 [00:00<00:01, 79994.41KB/s]
18%|#7 | 23768/132723 [00:00<00:01, 76680.01KB/s]
24%|##4 | 32330/132723 [00:00<00:01, 80117.08KB/s]
30%|### | 40361/132723 [00:00<00:01, 74549.15KB/s]
37%|###6 | 48855/132723 [00:00<00:01, 77872.28KB/s]
43%|####2 | 56707/132723 [00:00<00:01, 61594.84KB/s]
48%|####8 | 64310/132723 [00:00<00:01, 60647.30KB/s]
53%|#####3 | 70722/132723 [00:01<00:01, 51342.05KB/s]
60%|#####9 | 79226/132723 [00:01<00:00, 59245.91KB/s]
65%|######4 | 85680/132723 [00:01<00:00, 54381.58KB/s]
71%|####### | 94201/132723 [00:01<00:00, 61887.88KB/s]
76%|#######5 | 100855/132723 [00:01<00:00, 45686.07KB/s]
80%|######## | 106307/132723 [00:01<00:00, 43095.95KB/s]
86%|########6 | 114680/132723 [00:02<00:00, 43513.34KB/s]
92%|########
#1| 121550/132723 [00:02<00:00, 48665.68KB/s]
98%|#########8| 130086/132723 [00:02<00:00, 56963.55KB/s]
100%|##########| 132723/132723 [00:02<00:00, 58034.67KB/s]
@@ -246,7 +246,7 @@ Display result
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 4 minutes 3.870 seconds)
+ **Total running time of the script:** ( 4 minutes 4.065 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 5fbec72edc..3f4ff9e2bb 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
@@ -5,28 +5,28 @@
Computation times
=================
-**17:28.952** total execution time for **how_to_deploy_models** files:
+**17:11.643** total execution time for **how_to_deploy_models** files:
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 04:03.870 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``) | 04:04.065 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:52.109 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:41.966 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:32.983 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``) | 02:40.736 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:56.443 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``) | 01:52.802 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:21.891 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``) | 01:18.908 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 01:00.147 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``) | 00:55.492 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno_tvmc.py` (``deploy_model_on_adreno_tvmc.py``) | 00:56.226 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno_tvmc.py` (``deploy_model_on_adreno_tvmc.py``) | 00:54.722 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:46.295 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``) | 00:44.215 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:29.795 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``) | 00:29.579 | 0.0 MB |
+------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:29.187 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``) | 00:29.152 | 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 71483edb24..8228474c80 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.zip0c889e00-21d9-47ea-ad4b-37a33656e572 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+ Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip7a614ddb-3585-4d6c-bc26-b24d2b3b1ce4 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 aa22be8c35..75f645d58b 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
=================
-**00:57.089** total execution time for **how_to_extend_tvm** files:
+**00:55.739** 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``) | 00:52.825 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:51.813 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:03.131 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``) | 00:02.822 | 0.0 MB |
+-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.125 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``) | 00:01.097 | 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 76d07bbb67..e038546a19 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: 23251us [23251us] (48.99%; 48.99%)
- FoldScaleAxis: 24207us [9us] (51.01%; 51.01%)
- FoldConstant: 24198us [1733us] (50.99%; 99.96%)
- InferType: 22466us [22466us] (47.34%; 92.84%)
+ InferType: 22959us [22959us] (48.77%; 48.77%)
+ FoldScaleAxis: 24120us [7us] (51.23%; 51.23%)
+ FoldConstant: 24112us [1699us] (51.22%; 99.97%)
+ InferType: 22414us [22414us] (47.61%; 92.95%)
@@ -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: 22506us [22506us] (48.06%; 48.06%)
- FoldScaleAxis: 24319us [7us] (51.94%; 51.94%)
- FoldConstant: 24312us [1759us] (51.92%; 99.97%)
- InferType: 22553us [22553us] (48.16%; 92.76%)
+ InferType: 22696us [22696us] (48.29%; 48.29%)
+ FoldScaleAxis: 24300us [7us] (51.71%; 51.71%)
+ FoldConstant: 24294us [1739us] (51.69%; 99.97%)
+ InferType: 22554us [22554us] (47.99%; 92.84%)
diff --git a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
index 569aa1031c..98f5fb4a11 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: 53.514942 ms
+ Convolution: 38.739711 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 44da270e8e..e2321bc73b 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
@@ -598,7 +598,7 @@ be able to run on our build server
.. code-block:: none
- conv2d with tensor core: 12.233948 ms
+ conv2d with tensor core: 12.234931 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 4d2e08ad7d..41fc8e6041 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: 0.018487
- Baseline: 3.292571
+ Numpy running time: 0.017871
+ Baseline: 3.441978
@@ -227,7 +227,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
.. code-block:: none
- Opt1: 0.310587
+ Opt1: 0.300670
@@ -318,7 +318,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
.. code-block:: none
- Opt2: 0.343112
+ Opt2: 0.329739
@@ -406,7 +406,7 @@ the access pattern for A matrix is more cache friendly.
.. code-block:: none
- Opt3: 0.115167
+ Opt3: 0.114937
@@ -523,7 +523,7 @@ flattening.
.. code-block:: none
- Opt4: 0.109182
+ Opt4: 0.110421
@@ -635,7 +635,7 @@ write to C when all the block results are ready.
.. code-block:: none
- Opt5: 0.110476
+ Opt5: 0.114298
@@ -748,7 +748,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
.. code-block:: none
- Opt6: 0.146319
+ Opt6: 0.147723
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 982152e3ac..b9a0f54df7 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
=================
-**00:35.166** total execution time for **how_to_optimize_operators** files:
+**00:35.429** total execution time for **how_to_optimize_operators** files:
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.177 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``) | 00:32.450 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.862 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.904 | 0.0 MB |
+-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.127 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``) | 00:01.076 | 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 a4a2189f24..cd01c16fa2 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
=================
-**10:38.424** total execution time for **how_to_tune_with_autoscheduler** files:
+**10:41.295** 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``) | 06:38.700 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 06:27.997 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:44.702 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``) | 01:46.055 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:12.714 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``) | 01:12.916 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:33.748 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``) | 00:45.647 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.574 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``) | 00:14.723 | 0.0 MB |
+----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.986 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``) | 00:13.958 | 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 6f29beb2a1..2cb3d27a49 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
@@ -245,791 +245,518 @@ cooperative fetching, unrolling and operator fusion.
T.func_attr({"from_legacy_te_schedule": T.bool(True), "global_symbol": "main", "tir.noalias": T.bool(True)})
blockIdx_x = T.launch_thread("blockIdx.x", 32)
conv2d_nchw = T.allocate([14], "float32", "local")
- pad_temp_shared = T.allocate([6272], "float32", "shared")
- kernel_shared = T.allocate([2048], "float32", "shared")
+ pad_temp_shared = T.allocate([4032], "float32", "shared")
+ kernel_shared = T.allocate([3072], "float32", "shared")
threadIdx_x = T.launch_thread("threadIdx.x", 56)
- conv2d_nchw_1 = T.Buffer((4,), data=conv2d_nchw, scope="local", align=8)
+ conv2d_nchw_1 = T.Buffer((49,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
- conv2d_nchw_1[2] = T.float32(0)
- conv2d_nchw_1[4] = T.float32(0)
- conv2d_nchw_1[6] = T.float32(0)
- conv2d_nchw_1[8] = T.float32(0)
- conv2d_nchw_1[10] = T.float32(0)
- conv2d_nchw_1[12] = T.float32(0)
- conv2d_nchw_1[1] = T.float32(0)
- conv2d_nchw_1[3] = T.float32(0)
- conv2d_nchw_1[5] = T.float32(0)
conv2d_nchw_1[7] = T.float32(0)
+ conv2d_nchw_1[1] = T.float32(0)
+ conv2d_nchw_1[8] = T.float32(0)
+ conv2d_nchw_1[2] = T.float32(0)
conv2d_nchw_1[9] = T.float32(0)
+ conv2d_nchw_1[3] = T.float32(0)
+ conv2d_nchw_1[10] = T.float32(0)
+ conv2d_nchw_1[4] = T.float32(0)
conv2d_nchw_1[11] = T.float32(0)
+ conv2d_nchw_1[5] = T.float32(0)
+ conv2d_nchw_1[12] = T.float32(0)
+ conv2d_nchw_1[6] = T.float32(0)
conv2d_nchw_1[13] = T.float32(0)
- for rc_outer_outer, ry_outer_outer, rx_outer_outer in T.grid(4, 3, 3):
- cse_var_2: T.int32 = rc_outer_outer * 1152
- cse_var_1: T.int32 = ry_outer_outer * 3
+ for rc_outer_outer, rx_outer_outer in T.grid(8, 3):
+ cse_var_2: T.int32 = rc_outer_outer * 3136
+ cse_var_1: T.int32 = rc_outer_outer * 576
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((6272,), data=pad_temp_shared, scope="shared")
+ pad_temp_shared_1 = T.Buffer((4032,), data=pad_temp_shared, scope="shared")
data_1 = T.Buffer((25088,), data=data.data)
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 48], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 104], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 168] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 160], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 216], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 280] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 272], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 328], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 392] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 384], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 440], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 504] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 496], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 552], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 616] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 608], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 664], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 728] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 720], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 776], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 840] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 832], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 888], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 952] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 944], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1000], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1064] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1056], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1112], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1176] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1168], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1224], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1288] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1280], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1344] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1336], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1400] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1392], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1456] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1448], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1512] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1504], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1568] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1560], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1624] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1616], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1680] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1672], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1736] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1728], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 56) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1792] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1784], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 112) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1848] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1840], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 168] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 168) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1904] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1896], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1960] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1952], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 280] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 280) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2016] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2008], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 336) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2072] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2064], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 392] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 392) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2128] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2120], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 448) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2184] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2176], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 504] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 384], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2240] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2232], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 560) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2296] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2288], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 616] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 616) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2352] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2344], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 672) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2408] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2400], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 728] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 728) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2464] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2456], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 784) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2520] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2512], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 840] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 840) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2576] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2568], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 896) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2632] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2624], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 952] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 952) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2688] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2680], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 776], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2744] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2736], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1064] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1064) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2800] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2792], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1120) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2856] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2848], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1176] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1176) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2912] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2904], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1232) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2968] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2960], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1288] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1288) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3024] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3016], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1344] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1344) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3080] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3072], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1400] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1400) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3136] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3128], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1456] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1456) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3192] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3184], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1512] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1168], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3248] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3240], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1568] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1568) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3304] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3296], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1624] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1624) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3360] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3352], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1680] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1680) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3416] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3408], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1736] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1736) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3472] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3464], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1792] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1792) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3528] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3520], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1848] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1848) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3584] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3576], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1904] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1904) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3640] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3632], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1960] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1960) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3696] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3688], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2016] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1560], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3752] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3744], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2072] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2072) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3808] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3800], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2128] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2128) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3864] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3856], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2184] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2184) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3920] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3912], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2240] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2240) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3976] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3968], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2296] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2296) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4032] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4024], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2352] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2352) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4088] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4080], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2408] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2408) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4144] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4136], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2464] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2464) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4200] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4192], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2520] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1952], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4256] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4248], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2576] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2576) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4312] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4304], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2632] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2632) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4368] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4360], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2688] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2688) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4424] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4416], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2744] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2744) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4480] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4472], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2800] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2800) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4536] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4528], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2856] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2856) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4592] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4584], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2912] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2912) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4648] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4640], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2968] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2968) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4704] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4696], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3024] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 2344], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4760] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4752], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3080] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3080) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4816] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4808], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3136] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3136) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4872] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4864], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3192] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3192) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4928] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4920], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3248] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3248) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4984] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4976], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3304] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3304) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5040] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5032], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3360] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3360) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5096] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5088], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3416] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3416) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5152] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5144], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3472] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3472) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5208] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5200], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3528] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 2736], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5264] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5256], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3584] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3584) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5320] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5312], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3640] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3640) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5376] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5368], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3696] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3696) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5432] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5424], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3752] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3752) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5488] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5480], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3808] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3808) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5544] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5536], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3864] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3864) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5600] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5592], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3920] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3920) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5656] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5648], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5712] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5704], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5768] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5760], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5824] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5816], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5880] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5872], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5936] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5928], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5992] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5984], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6048] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6040], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6104] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6096], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6160] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6152], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6216] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6208], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3976] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3976) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
- kernel_shared_1 = T.Buffer((2048,), data=kernel_shared, scope="shared")
+ kernel_shared_1 = T.Buffer((3072,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 168) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 224) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 280) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 336) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 144) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 392) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 24]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 448) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 192]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 504) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 360]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 560) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 176) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 616) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 120]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 672) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 288]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 728) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 152) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 784) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 48]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 840) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 216]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 112) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 896) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 384]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 168) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 952) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 184) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 224) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 96) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1008) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 144]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 280) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 216]
+ kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1064) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 312]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 336) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 80) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1120) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 160) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 392) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 72]
+ kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1176) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 72]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 448) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 576]
+ kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1232) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 240]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 504) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 120) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1288) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 408]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 560) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 432]
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 32256]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 616) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 104) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1400) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 672) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 288]
+ kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1456) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 728) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 88) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1512) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 784) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 144]
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1568) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 840) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 648]
+ kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1624) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 32256]
+ kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1680) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 144) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 952) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1736) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 24]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1008) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1792) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 192]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1064) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1848) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 360]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1120) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 96) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1904) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 176) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1176) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 216]
+ kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1960) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 120]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1232) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 80) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2016) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 288]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1288) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 72]
+ kernel_shared_1[threadIdx_x_2 + 2072] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2072) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 152) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1344) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 576]
+ kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2128) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 48]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1400) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 120) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2184] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2184) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 216]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1456) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 432]
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2240) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 384]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1512) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 104) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2296] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2296) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 184) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1568) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 288]
+ kernel_shared_1[threadIdx_x_2 + 2352] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2352) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 144]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1624) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 88) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2408] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2408) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 312]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1680) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 144]
+ kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2464) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 160) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1736) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 648]
+ kernel_shared_1[threadIdx_x_2 + 2520] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2520) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 72]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 64512]
+ kernel_shared_1[threadIdx_x_2 + 2576] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2576) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 240]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1848) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 2632] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2632) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 408]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1904) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 64512]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1960) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 2744] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2744) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
with T.launch_thread(threadIdx_x_2, 56):
- if T.likely(threadIdx_x_2 < 32):
- kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2016) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 864]
- for rc_outer_inner in range(4):
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 49] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 50] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 51] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 52] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 53] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 49] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 50] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 51] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 52] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 53] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 98] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 102] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 103] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 104] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 98] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 102] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 103] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 104] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 147] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 148] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 149] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 150] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 151] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 152] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 153] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 147] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 148] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 149] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 150] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 151] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 152] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 153] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 196] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 197] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 198] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 199] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 200] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 201] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 202] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 196] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 197] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 198] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 199] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 200] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 201] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 202] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 245] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 246] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 247] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 248] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 249] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 250] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 251] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 245] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 246] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 247] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 248] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 249] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 250] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 251] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 294] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 295] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 296] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 297] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 298] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 299] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 300] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 294] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 295] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 296] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 297] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 298] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 299] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 300] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 343] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 344] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 345] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 346] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 347] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 348] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 349] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 343] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 344] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 345] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 346] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 347] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 348] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 349] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 392] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 393] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 394] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 395] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 396] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 397] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 398] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 392] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 393] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 394] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 395] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 396] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 397] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 398] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 490] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 491] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 492] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 493] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 494] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 495] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 496] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 490] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 491] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 492] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 493] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 494] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 495] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 496] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 539] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 540] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 541] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 542] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 543] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 544] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 545] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 539] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 540] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 541] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 542] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 543] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 544] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 545] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 588] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 589] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 590] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 591] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 592] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 593] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 594] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 588] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 589] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 590] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 591] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 592] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 593] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 594] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 637] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 638] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 639] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 640] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 641] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 642] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 643] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 637] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 638] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 639] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 640] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 641] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 642] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 643] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 686] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 687] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 688] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 689] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 690] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 691] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 692] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 686] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 687] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 688] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 689] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 690] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 691] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 692] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 735] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 736] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 737] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 738] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 739] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 740] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 741] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 735] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 736] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 737] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 738] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 739] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 740] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 741] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 784] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 785] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 786] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 787] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 788] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 789] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 790] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 784] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 785] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 786] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 787] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 788] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 789] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 790] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 833] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 834] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 835] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 836] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 837] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 838] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 839] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 833] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 834] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 835] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 836] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 837] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 838] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 839] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 931] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 932] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 933] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 934] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 935] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 936] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 937] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 931] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 932] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 933] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 934] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 935] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 936] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 937] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 980] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 981] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 982] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 983] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 984] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 985] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 986] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 980] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 981] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 982] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 983] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 984] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 985] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 986] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1029] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1030] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1031] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1032] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1033] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1034] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1035] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1029] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1030] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1031] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1032] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1033] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1034] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1035] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1078] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1079] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1080] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1081] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1082] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1083] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1084] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1078] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1079] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1080] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1081] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1082] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1083] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1084] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1127] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1128] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1129] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1130] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1131] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1132] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1133] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1127] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1128] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1129] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1130] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1131] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1132] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1133] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1176] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1177] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1178] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1179] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1180] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1181] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1182] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1176] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1177] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1178] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1179] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1180] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1181] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1182] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1225] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1226] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1227] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1228] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1229] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1230] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1231] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1225] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1226] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1227] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1228] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1229] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1230] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1231] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1274] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1275] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1276] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1277] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1278] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1279] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1280] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1274] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1275] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1276] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1277] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1278] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1279] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1280] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1323] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1324] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1325] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1326] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1327] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1328] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1329] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1323] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1324] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1325] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1326] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1327] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1328] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1329] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1372] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1373] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1374] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1375] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1376] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1377] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1378] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1372] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1373] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1374] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1375] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1376] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1377] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1378] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1421] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1422] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1423] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1424] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1425] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1426] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1427] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1421] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1422] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1423] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1424] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1425] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1426] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1427] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1470] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1471] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1472] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1473] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1474] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1475] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1476] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1470] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1471] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1472] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1473] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1474] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1475] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1476] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1519] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1520] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1521] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1522] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1523] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1524] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1525] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1519] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1520] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1521] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1522] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1523] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1524] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1525] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- for i1_inner in range(2):
+ kernel_shared_1[threadIdx_x_2 + 2800] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2800) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2856] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2856) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2912) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2968] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2968) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
+ with T.launch_thread(threadIdx_x_2, 56):
+ if T.likely(threadIdx_x_2 < 48):
+ kernel_shared_1[threadIdx_x_2 + 3024] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 3024) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 432]
+ for rc_outer_inner, ry_outer_inner in T.grid(4, 3):
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 126] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 126] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 189] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 189] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 252] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 252] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 315] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 315] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 378] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 378] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 504] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 504] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 567] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 567] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 630] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 630] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 693] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 693] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 756] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 756] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 819] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 819] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 945] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 945] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 127] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 127] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 190] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 190] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 253] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 253] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 316] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 316] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 379] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 379] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 505] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 505] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 568] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 568] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 631] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 631] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 694] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 694] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 757] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 757] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 820] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 820] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 946] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 946] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 128] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 128] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 191] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 191] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 254] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 254] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 317] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 317] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 380] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 380] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 506] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 506] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 569] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 569] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 632] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 632] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 695] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 695] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 758] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 758] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 821] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 821] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 947] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 947] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 66] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 66] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 129] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 129] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 192] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 192] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 255] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 255] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 318] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 318] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 381] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 381] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 507] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 507] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 570] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 570] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 633] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 633] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 696] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 696] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 759] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 759] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 822] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 822] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 948] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 948] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 67] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 67] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 130] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 130] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 193] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 193] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 256] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 256] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 319] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 319] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 382] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 382] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 508] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 508] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 571] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 571] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 634] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 634] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 697] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 697] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 760] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 760] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 823] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 823] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 949] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 949] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 68] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 68] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 131] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 131] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 194] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 194] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 257] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 257] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 320] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 320] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 383] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 383] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 509] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 509] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 572] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 572] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 635] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 635] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 698] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 698] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 761] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 761] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 824] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 824] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 950] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 950] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 69] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 69] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 132] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 132] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 195] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 195] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 258] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 258] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 321] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 321] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 384] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 384] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 510] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 510] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 573] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 573] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 636] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 636] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 699] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 699] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 762] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 762] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 825] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 825] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 951] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 951] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ 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 * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 1] = T.max(conv2d_nchw_1[i1_inner + 2] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 2] = T.max(conv2d_nchw_1[i1_inner + 4] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 3] = T.max(conv2d_nchw_1[i1_inner + 6] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 4] = T.max(conv2d_nchw_1[i1_inner + 8] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 5] = T.max(conv2d_nchw_1[i1_inner + 10] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 6] = T.max(conv2d_nchw_1[i1_inner + 12] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x * 784 + threadIdx_x * 7 + i3_inner] = T.max(conv2d_nchw_1[i3_inner] + bias_1[blockIdx_x * 16 + threadIdx_x // 7], T.float32(0))
+ compute_1[blockIdx_x * 784 + threadIdx_x * 7 + i3_inner + 392] = T.max(conv2d_nchw_1[i3_inner + 7] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 + 8], T.float32(0))
@@ -1079,7 +806,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 0.257 ms
+ Execution time of this operator: 0.230 ms
@@ -1127,37 +854,37 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
- conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
+ conv2d_nchw_ff_o_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=8)
- conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+ conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
- conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
+ conv2d_nchw_xx_o_o_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=7)
- conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=32)
+ 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=16)
conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
- conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
+ conv2d_nchw_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=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=2)
+ 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=8)
- compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+ compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
- compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
+ compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
- compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=7)
+ compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
kernel_shared = s.cache_read(kernel, "shared", [conv2d_nchw])
@@ -1183,7 +910,7 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
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=56)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
- s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
+ s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -1210,639 +937,388 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
#endif
extern "C" __global__ void __launch_bounds__(56) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
float conv2d_nchw[14];
- __shared__ float pad_temp_shared[6272];
- __shared__ float kernel_shared[2048];
+ __shared__ float pad_temp_shared[4032];
+ __shared__ float kernel_shared[3072];
conv2d_nchw[0] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
- conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
- conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[1] = 0.000000e+00f;
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[5] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[13] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 4; ++rc_outer_outer) {
- for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
- for (int rx_outer_outer = 0; rx_outer_outer < 3; ++rx_outer_outer) {
- __syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 48)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 104)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 168)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 160)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 216)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 280)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 272)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 328)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 392)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 384)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 448)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 440)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 504)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 496)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 552)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 616)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 608)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 664)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 728)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 720)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 776)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 840)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 832)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 888)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 952)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 944)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1008)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1000)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1064)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1056)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1112)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1176)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1168)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1232)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1224)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1288)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1280)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1344)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1336)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1400)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1392)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1456)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1448)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1512)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1504)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1568)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1560)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1624)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1616)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1680)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1672)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1736)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1728)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1792)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1784)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1848)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1840)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1904)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1896)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1960)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1952)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2016)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2008)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2072)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2064)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2128)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2120)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2184)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2176)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2240)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2232)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2296)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2288)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2352)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2344)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2408)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2400)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2464)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2456)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2520)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2512)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2576)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2568)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2632)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2624)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2688)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2680)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2744)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2736)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2800)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2792)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2856)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2848)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2912)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2904)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2968)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2960)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3024)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3016)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3080)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3072)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3136)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3128)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3192)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3184)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3248)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3240)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3304)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3296)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3360)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3352)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3416)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3408)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3472)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3464)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3528)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3520)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3584)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3576)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3640)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3632)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3696)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3688)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3752)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3744)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3808)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3800)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3864)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3856)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3920)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3912)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3976)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3968)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4032)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4024)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4088)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4080)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4144)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4136)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4200)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4192)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4256)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4248)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4312)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4304)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4368)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4360)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4424)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4416)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4480)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4472)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4536)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4528)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4592)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4584)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4648)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4640)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4704)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4696)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4760)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4752)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4816)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4808)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4872)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4864)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4928)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4920)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4984)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4976)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5040)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5032)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5096)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5088)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5152)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5144)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5208)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5200)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5264)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5256)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5320)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5312)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5376)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5368)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5432)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5424)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5488)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5480)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5544)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5536)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5600)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5592)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5656)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5648)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5712)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5704)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5768)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5760)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5824)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5816)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5880)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5872)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5936)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5928)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5992)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5984)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6048)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6040)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6104)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6096)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6160)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6152)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6216)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6208)] : 0.000000e+00f);
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 56)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 112)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 112) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 168) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 224) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 96) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 280)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 280) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 216)];
- kernel_shared[(((int)threadIdx.x) + 336)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 336) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 80) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 392)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 392) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 72)];
- kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 448) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 576)];
- kernel_shared[(((int)threadIdx.x) + 504)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 504) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 120) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 560) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 432)];
- kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 616) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 104) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 672) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 288)];
- kernel_shared[(((int)threadIdx.x) + 728)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 728) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 88) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 784) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 144)];
- kernel_shared[(((int)threadIdx.x) + 840)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 840) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 648)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 32256)];
- kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 952) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1008) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1064) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1120) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 96) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1176) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 216)];
- kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1232) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 80) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1288) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 72)];
- kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1344) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 576)];
- kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1400) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 120) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1456) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 432)];
- kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1512) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 104) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1568) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 288)];
- kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1624) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 88) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1680) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 144)];
- kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1736) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 648)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 64512)];
- kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1848) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1904) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1960) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- if (((int)threadIdx.x) < 32) {
- kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2016) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 864)];
- }
- __syncthreads();
- for (int rc_outer_inner = 0; rc_outer_inner < 4; ++rc_outer_inner) {
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 49)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 50)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 51)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 52)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 53)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 54)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 55)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 49)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 50)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 51)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 52)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 53)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 54)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 55)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 98)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 99)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 100)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 101)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 102)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 103)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 104)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 98)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 99)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 100)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 101)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 102)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 103)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 104)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 147)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 148)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 149)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 150)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 151)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 152)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 153)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 147)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 148)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 149)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 150)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 151)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 152)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 153)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 196)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 197)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 198)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 199)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 200)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 201)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 202)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 196)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 197)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 198)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 199)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 200)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 201)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 202)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 245)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 246)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 247)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 248)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 249)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 250)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 251)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 245)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 246)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 247)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 248)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 249)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 250)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 251)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 294)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 295)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 296)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 297)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 298)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 299)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 300)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 294)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 295)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 296)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 297)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 298)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 299)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 300)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 343)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 344)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 345)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 346)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 347)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 348)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 349)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 343)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 344)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 345)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 346)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 347)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 348)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 349)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 392)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 393)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 394)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 395)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 396)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 397)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 398)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 392)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 393)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 394)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 395)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 396)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 397)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 398)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 490)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 491)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 492)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 493)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 494)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 495)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 496)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 490)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 491)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 492)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 493)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 494)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 495)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 496)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 539)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 540)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 541)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 542)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 543)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 544)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 545)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 539)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 540)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 541)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 542)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 543)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 544)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 545)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 588)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 589)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 590)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 591)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 592)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 593)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 594)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 588)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 589)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 590)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 591)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 592)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 593)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 594)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 637)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 638)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 639)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 640)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 641)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 642)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 643)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 637)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 638)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 639)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 640)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 641)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 642)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 643)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 686)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 687)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 688)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 689)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 690)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 691)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 692)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 686)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 687)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 688)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 689)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 690)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 691)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 692)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 735)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 736)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 737)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 738)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 739)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 740)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 741)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 735)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 736)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 737)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 738)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 739)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 740)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 741)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 784)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 785)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 786)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 787)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 788)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 789)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 790)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 784)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 785)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 786)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 787)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 788)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 789)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 790)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 833)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 834)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 835)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 836)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 837)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 838)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 839)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 833)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 834)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 835)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 836)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 837)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 838)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 839)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 931)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 932)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 933)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 934)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 935)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 936)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 937)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 931)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 932)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 933)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 934)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 935)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 936)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 937)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 980)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 981)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 982)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 983)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 984)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 985)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 986)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 980)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 981)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 982)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 983)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 984)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 985)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 986)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1029)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1030)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1031)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1032)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1033)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1034)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1035)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1029)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1030)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1031)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1032)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1033)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1034)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1035)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1078)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1079)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1080)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1081)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1082)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1083)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1084)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1078)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1079)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1080)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1081)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1082)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1083)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1084)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1127)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1128)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1129)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1130)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1131)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1132)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1133)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1127)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1128)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1129)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1130)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1131)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1132)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1133)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1176)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1177)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1178)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1179)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1180)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1181)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1182)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1176)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1177)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1178)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1179)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1180)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1181)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1182)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1225)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1226)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1227)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1228)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1229)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1230)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1231)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1225)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1226)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1227)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1228)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1229)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1230)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1231)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1274)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1275)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1276)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1277)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1278)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1279)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1280)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1274)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1275)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1276)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1277)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1278)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1279)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1280)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1323)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1324)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1325)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1326)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1327)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1328)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1329)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1323)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1324)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1325)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1326)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1327)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1328)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1329)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1372)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1373)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1374)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1375)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1376)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1377)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1378)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1372)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1373)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1374)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1375)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1376)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1377)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1378)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1421)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1422)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1423)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1424)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1425)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1426)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1427)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1421)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1422)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1423)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1424)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1425)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1426)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1427)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1470)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1471)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1472)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1473)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1474)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1475)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1476)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1470)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1471)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1472)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1473)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1474)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1475)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1476)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1519)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1520)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1521)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1522)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1523)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1524)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1525)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1519)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1520)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1521)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1522)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1523)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1524)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1525)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
+ for (int rc_outer_outer = 0; rc_outer_outer < 8; ++rc_outer_outer) {
+ for (int rx_outer_outer = 0; rx_outer_outer < 3; ++rx_outer_outer) {
+ __syncthreads();
+ pad_temp_shared[((int)threadIdx.x)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 56) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 112) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 168)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 168) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 224) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 280)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 280) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 336) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 392)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 392) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 448)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 448) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 504)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 384)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 560) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 616)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 616) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 672) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 728)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 728) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 784) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 840)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 840) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 896) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 952)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 952) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1008)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 776)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1064)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1064) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1120) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1176)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1176) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1232)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1232) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1288)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1288) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1344)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1344) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1400)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1400) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1456)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1456) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1512)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1168)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1568)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1568) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1624)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1624) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1680)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1680) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1736)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1736) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1792)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1792) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1848)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1848) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1904)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1904) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1960)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1960) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2016)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1560)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2072)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2072) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2128)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2128) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2184)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2184) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2240)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2240) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2296)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2296) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2352)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2352) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2408)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2408) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2464)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2464) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2520)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1952)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2576)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2576) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2632)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2632) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2688)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2688) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2744)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2744) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2800)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2800) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2856)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2856) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2912)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2912) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2968)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2968) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3024)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 2344)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3080)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3080) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3136)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3136) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3192)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3192) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3248)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3248) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3304)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3304) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3360)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3360) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3416)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3416) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3472)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3472) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3528)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 2736)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3584)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3584) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3640)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3640) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3696)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3696) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3752)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3752) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3808)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3808) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3864)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3864) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3920)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3920) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3976)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3976) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 56)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 112)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 168) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 224) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 280)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 280) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 336) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 144) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 392)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 392) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 24)];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 448) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 192)];
+ kernel_shared[(((int)threadIdx.x) + 504)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 504) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 360)];
+ kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 560) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 176) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 616) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 120)];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 672) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 288)];
+ kernel_shared[(((int)threadIdx.x) + 728)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 728) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 152) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 784)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 784) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 840)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 840) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 216)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 896) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 384)];
+ kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 952) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 184) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1008) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 144)];
+ kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1064) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 312)];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1120) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 160) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1176) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 72)];
+ kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1232) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 240)];
+ kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1288) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 408)];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 32256)];
+ kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1400) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1456) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1512) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1568) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1624) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1680) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 144) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1736) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 24)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1792) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 192)];
+ kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1848) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 360)];
+ kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1904) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 176) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1960) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 120)];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2016) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 288)];
+ kernel_shared[(((int)threadIdx.x) + 2072)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2072) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 152) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2128) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2184)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2184) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 216)];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2240) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 384)];
+ kernel_shared[(((int)threadIdx.x) + 2296)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2296) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 184) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2352)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2352) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 144)];
+ kernel_shared[(((int)threadIdx.x) + 2408)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2408) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 312)];
+ kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2464) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 160) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2520)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2520) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 72)];
+ kernel_shared[(((int)threadIdx.x) + 2576)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2576) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 240)];
+ kernel_shared[(((int)threadIdx.x) + 2632)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2632) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 408)];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 2744)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2744) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 2800)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2800) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 2856)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2856) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2912) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2968)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2968) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ if (((int)threadIdx.x) < 48) {
+ kernel_shared[(((int)threadIdx.x) + 3024)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 3024) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 432)];
+ }
+ __syncthreads();
+ for (int rc_outer_inner = 0; rc_outer_inner < 4; ++rc_outer_inner) {
+ for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 63)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 63)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 126)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 126)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 189)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 189)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 252)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 252)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 315)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 315)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 378)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 378)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 504)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 504)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 567)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 567)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 630)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 630)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 693)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 693)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 756)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 756)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 819)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 819)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 945)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 945)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 64)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 64)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 127)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 127)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 190)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 190)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 253)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 253)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 316)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 316)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 379)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 379)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 505)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 505)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 568)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 568)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 631)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 631)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 694)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 694)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 757)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 757)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 820)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 820)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 946)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 946)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 65)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 65)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 128)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 128)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 191)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 191)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 254)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 254)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 317)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 317)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 380)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 380)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 506)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 506)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 569)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 569)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 632)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 632)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 695)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 695)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 758)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 758)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 821)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 821)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 947)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 947)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 66)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 66)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 129)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 129)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 192)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 192)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 255)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 255)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 318)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 318)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 381)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 381)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 507)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 507)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 570)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 570)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 633)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 633)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 696)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 696)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 759)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 759)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 822)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 822)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 948)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 948)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 67)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 67)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 130)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 130)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 193)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 193)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 256)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 256)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 319)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 319)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 382)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 382)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 508)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 508)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 571)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 571)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 634)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 634)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 697)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 697)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 760)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 760)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 823)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 823)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 949)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 949)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 68)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 68)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 131)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 131)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 194)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 194)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 257)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 257)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 320)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 320)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 383)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 383)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 509)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 509)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 572)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 572)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 635)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 635)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 698)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 698)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 761)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 761)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 824)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 824)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 950)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 950)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 69)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 69)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 132)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 132)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 195)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 195)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 258)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 258)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 321)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 321)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 384)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 384)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 510)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 510)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 573)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 573)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 636)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 636)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 699)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 699)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 762)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 762)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 825)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 825)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 951)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 951)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
}
}
}
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7))] = max((conv2d_nchw[i1_inner] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] = max((conv2d_nchw[(i1_inner + 2)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] = max((conv2d_nchw[(i1_inner + 6)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] = max((conv2d_nchw[(i1_inner + 10)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[(((((int)blockIdx.x) * 784) + (((int)threadIdx.x) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 7))]), 0.000000e+00f);
+ compute[((((((int)blockIdx.x) * 784) + (((int)threadIdx.x) * 7)) + i3_inner) + 392)] = max((conv2d_nchw[(i3_inner + 7)] + bias[(((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 7)) + 8)]), 0.000000e+00f);
}
}
@@ -1902,7 +1378,7 @@ In the example below we resume the status and do more 5 trials.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 6 minutes 38.700 seconds)
+ **Total running time of the script:** ( 6 minutes 27.997 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 3a19d8a4e4..2abc517dc8 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)
- 8.1015 8.1062 8.1119 8.0863 0.0110
+ 8.0760 8.0819 8.0826 8.0636 0.0088
@@ -675,7 +675,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 12.714 seconds)
+ **Total running time of the script:** ( 1 minutes 12.916 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 4db4bce9fb..d844553e57 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)
- 752.6929 752.2915 754.9909 750.7964 1.7358
+ 752.5467 752.2150 753.9007 751.5244 0.9981
@@ -694,7 +694,7 @@ Other Tips
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 44.702 seconds)
+ **Total running time of the script:** ( 1 minutes 46.055 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 cdd68b6b7a..fa8aaae24e 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
@@ -389,27 +389,27 @@ 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": T.bool(True), "global_symbol": "main", "tir.noalias": T.bool(True)})
- for i0_outer in T.parallel(4):
- compute_1 = T.allocate([1024], "float32", "global")
- for i1_outer in range(16):
- compute_2 = T.Buffer((1024,), data=compute_1)
- for i_outer_inner, nb_j_inner in T.grid(2, 2):
- for i_inner_init, j_init in T.grid(16, 16):
- compute_2[i_outer_inner * 512 + i_inner_init * 32 + nb_j_inner * 16 + j_init] = T.float32(0)
- for elem_idx, i_inner, j in T.grid(T.Let(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1], where={cse_var_1: i1_outer * 2 + nb_j_inner}), 16, 16):
- cse_var_1 = T.int32()
- placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- cse_var_3: T.int32 = i1_outer * 2 + nb_j_inner
- cse_var_2: T.int32 = i_outer_inner * 512 + i_inner * 32 + nb_j_inner * 16 + j
+ for i0_outer_i1_outer_fused in T.parallel(512):
+ compute_1 = T.allocate([128], "float32", "global")
+ compute_2 = T.Buffer((128,), data=compute_1)
+ for i_outer_inner in range(2):
+ for i_inner_init, j_init in T.grid(4, 16):
+ compute_2[i_outer_inner * 64 + i_inner_init * 16 + j_init] = T.float32(0)
+ for elem_idx, i_inner, j in T.grid(T.Let(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1], where={cse_var_1: i0_outer_i1_outer_fused % 32}), 4, 16):
+ cse_var_1 = T.int32()
+ placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
+ cse_var_2: T.int32 = i0_outer_i1_outer_fused % 32
+ if T.likely(elem_idx < placeholder_5[cse_var_2 + 1] - placeholder_5[cse_var_2]):
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 * 8192 + i_outer_inner * 4096 + 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 * 16384 + i0_inner * 512 + i1_outer * 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))
+ cse_var_3: T.int32 = i_outer_inner * 64 + i_inner * 16 + j
+ compute_2[cse_var_3] = compute_2[cse_var_3] + placeholder_6[placeholder_5[cse_var_2] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + i_outer_inner * 1024 + i_inner * 256 + placeholder_8[placeholder_5[cse_var_2] + elem_idx]], T.float32(0))
+ for i0_inner in range(8):
+ cse_var_4: T.int32 = i0_outer_i1_outer_fused // 32 * 4096 + i0_inner * 512 + i0_outer_i1_outer_fused % 32 * 16
+ 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 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_4:cse_var_4 + 16], T.Broadcast(T.float32(0), 16))
@@ -459,7 +459,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 1.521 ms
+ Execution time of this operator: 1.336 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 bf993f884d..58e3ba1b9c 100644
--- a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
Computation times
=================
-**00:55.789** total execution time for **how_to_tune_with_autotvm** files:
+**00:37.052** 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``) | 00:55.753 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``) | 00:37.017 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``) | 00:00.020 | 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 970d59373c..a9f2ebb2d8 100644
--- a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
@@ -390,7 +390,7 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 128, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5738367
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2605244
No: 2 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -513,7 +513,7 @@ 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, 4, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1151832
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 64, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4000086
No: 3 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -636,9 +636,8 @@ 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, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4212501
- No: 4 GFLOPS: 1.02/1.02 result: MeasureResult(costs=(0.22650778175,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.790705442428589, timestamp=1682987906.4804852) [('tile_f', [-1, 8, 1, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3360248
- No: 5 GFLOPS: 0.00/1.02 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,533453
+ No: 4 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -760,8 +759,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8662698
- No: 6 GFLOPS: 0.00/1.02 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9991495
+ No: 5 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -883,11 +882,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8824596
- No: 7 GFLOPS: 30.20/30.20 result: MeasureResult(costs=(0.007664544214285715,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.030125617980957, timestamp=1682987910.191251) [('tile_f', [-1, 1, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4294570
- No: 8 GFLOPS: 161.19/161.19 result: MeasureResult(costs=(0.0014362217428571429,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2537076473236084, timestamp=1682987911.046372) [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5295007
- No: 9 GFLOPS: 28.82/161.19 result: MeasureResult(costs=(0.0080316095,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9334707260131836, timestamp=1682987922.23613) [('tile_f', [-1, 1, 1, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6271065
- No: 10 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6152202
+ No: 6 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1009,26 +1005,8 @@ 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, 16, 8, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5508265
- No: 11 GFLOPS: 0.00/161.19 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, 32, 2, 1]), ('tile_y', [-1, 7, 1, 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', 1)],None,9297435
- No: 12 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 1, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,188704
+ No: 7 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1150,8 +1128,8 @@ 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, 128, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7835177
- No: 13 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 1, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8251241
+ No: 8 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1273,8 +1251,8 @@ 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, 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', 1500), ('unroll_explicit', 0)],None,4548917
- No: 14 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 128, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3263751
+ No: 9 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1396,8 +1374,9 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5217649
- No: 15 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9042267
+ No: 10 GFLOPS: 40.99/40.99 result: MeasureResult(costs=(0.005647334888888888,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.6838254928588867, timestamp=1683002350.2191744) [('tile_f', [-1, 1, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2784230
+ No: 11 GFLOPS: 0.00/40.99 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
@@ -1519,9 +1498,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 1, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7311838
- No: 16 GFLOPS: 7.14/161.19 result: MeasureResult(costs=(0.03242240125,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.5403568744659424, timestamp=1682987927.5691445) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7716502
- No: 17 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10048353
+ No: 12 GFLOPS: 0.00/40.99 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
@@ -1643,8 +1621,776 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 128]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4791154
- No: 18 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 2, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1152285
+ No: 13 GFLOPS: 0.00/40.99 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 742, in __call__
+ yield remote, remote.load_module(os.path.split(build_result.filename)[1])
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
+ costs = time_f(*args).results
+ File "/workspace/python/tvm/runtime/module.py", line 399, in evaluator
+ blob = feval(*args)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 262, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 251, in tvm._ffi._cy3.core.FuncCall3
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+ tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 4: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 3: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 2: tvm::runtime::RPCWrappedFunc::operator()(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../src/runtime/rpc/rpc_module.cc:129
+ 1: tvm::runtime::RPCClientSession::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)> const&)
+ at ../src/runtime/rpc/rpc_endpoint.cc:1012
+ 0: tvm::runtime::RPCEndpoint::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)>)
+ at ../src/runtime/rpc/rpc_endpoint.cc:804
+ File "../src/runtime/rpc/rpc_endpoint.cc", line 804
+ TVMError:
+ ---------------------------------------------------------------
+ An error occurred during the execution of TVM.
+ For more information, please see: https://tvm.apache.org/docs/errors.html
+ ---------------------------------------------------------------
+ Check failed: (code == RPCCode::kReturn) is false: code=kShutdown
+
+ During handling of the above exception, another exception occurred:
+
+ Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
+ costs = time_f(*args).results
+ File "/usr/lib/python3.7/contextlib.py", line 130, in __exit__
+ self.gen.throw(type, value, traceback)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 746, in __call__
+ remote.remove(build_result.filename)
+ File "/workspace/python/tvm/rpc/client.py", line 144, in remove
+ self._remote_funcs["remove"] = self.get_function("tvm.rpc.server.remove")
+ File "/workspace/python/tvm/rpc/client.py", line 72, in get_function
+ return self._sess.get_function(name)
+ File "/workspace/python/tvm/runtime/module.py", line 179, in get_function
+ self.handle, c_str(name), ctypes.c_int(query_imports), ctypes.byref(ret_handle)
+ File "/workspace/python/tvm/_ffi/base.py", line 348, in check_call
+ raise get_last_ffi_error()
+ tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 52: 0xffffffffffffffff
+ 51: _start
+ 50: __libc_start_main
+ 49: _Py_UnixMain
+ 48: 0x0000000000650da0
+ 47: 0x0000000000650afa
+ 46: _PyFunction_FastCallDict
+ 45: _PyEval_EvalCodeWithName
+ 44: _PyEval_EvalFrameDefault
+ 43: _PyFunction_FastCallKeywords
+ 42: _PyEval_EvalCodeWithName
+ 41: _PyEval_EvalFrameDefault
+ 40: _PyMethodDef_RawFastCallKeywords
+ 39: 0x0000000000546369
+ 38: _PyEval_EvalCodeWithName
+ 37: _PyEval_EvalFrameDefault
+ 36: _PyFunction_FastCallKeywords
+ 35: _PyEval_EvalCodeWithName
+ 34: _PyEval_EvalFrameDefault
+ 33: _PyFunction_FastCallDict
+ 32: _PyEval_EvalCodeWithName
+ 31: _PyEval_EvalFrameDefault
+ 30: _PyObject_FastCallDict
+ 29: 0x00000000004c06e1
+ 28: _PyFunction_FastCallDict
+ 27: _PyEval_EvalFrameDefault
+ 26: _PyMethodDescr_FastCallKeywords
+ 25: 0x00000000005dcb58
+ 24: 0x00000000005dc83f
+ 23: 0x00000000004ba127
+ 22: _PyEval_EvalFrameDefault
+ 21: _PyFunction_FastCallKeywords
+ 20: _PyEval_EvalFrameDefault
+ 19: _PyFunction_FastCallKeywords
+ 18: _PyEval_EvalFrameDefault
+ 17: _PyFunction_FastCallKeywords
+ 16: _PyEval_EvalCodeWithName
+ 15: _PyEval_EvalFrameDefault
+ 14: 0x0000000000537c30
+ 13: _PyObject_FastCallKeywords
+ 12: 0x00007fd78302dfa2
+ 11: _ctypes_callproc
+ 10: ffi_call
+ 9: ffi_call_unix64
+ 8: TVMModGetFunction
+ at ../src/runtime/c_runtime_api.cc:408
+ 7: tvm::runtime::ModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, bool)
+ at ../src/runtime/module.cc:66
+ 6: tvm::runtime::RPCModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, tvm::runtime::ObjectPtr<tvm::runtime::Object> const&)
+ at ../src/runtime/rpc/rpc_module.cc:187
+ 5: tvm::runtime::RPCClientSession::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
+ at ../src/runtime/rpc/rpc_endpoint.cc:1007
+ 4: tvm::runtime::TVMRetValue tvm::runtime::RPCEndpoint::SysCallRemote<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(tvm::runtime::RPCCode, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
+ at ../src/runtime/rpc/rpc_endpoint.h:223
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(int&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) const
+ at ../include/tvm/runtime/packed_func.h:1621
+ 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/rpc/rpc_endpoint.cc:684
+ File "../src/runtime/rpc/rpc_endpoint.cc", line 684
+ TVMError:
+ ---------------------------------------------------------------
+ An error occurred during the execution of TVM.
+ For more information, please see: https://tvm.apache.org/docs/errors.html
+ ---------------------------------------------------------------
+ Check failed: (code == RPCCode::kReturn) is false: code=1
+
+ Traceback (most recent call last):
+ 52: 0xffffffffffffffff
+ 51: _start
+ 50: __libc_start_main
+ 49: _Py_UnixMain
+ 48: 0x0000000000650da0
+ 47: 0x0000000000650afa
+ 46: _PyFunction_FastCallDict
+ 45: _PyEval_EvalCodeWithName
+ 44: _PyEval_EvalFrameDefault
+ 43: _PyFunction_FastCallKeywords
+ 42: _PyEval_EvalCodeWithName
+ 41: _PyEval_EvalFrameDefault
+ 40: _PyMethodDef_RawFastCallKeywords
+ 39: 0x0000000000546369
+ 38: _PyEval_EvalCodeWithName
+ 37: _PyEval_EvalFrameDefault
+ 36: _PyFunction_FastCallKeywords
+ 35: _PyEval_EvalCodeWithName
+ 34: _PyEval_EvalFrameDefault
+ 33: _PyFunction_FastCallDict
+ 32: _PyEval_EvalCodeWithName
+ 31: _PyEval_EvalFrameDefault
+ 30: _PyObject_FastCallDict
+ 29: 0x00000000004c06e1
+ 28: _PyFunction_FastCallDict
+ 27: _PyEval_EvalFrameDefault
+ 26: _PyMethodDescr_FastCallKeywords
+ 25: 0x00000000005dcb58
+ 24: 0x00000000005dc83f
+ 23: 0x00000000004ba127
+ 22: _PyEval_EvalFrameDefault
+ 21: _PyFunction_FastCallKeywords
+ 20: _PyEval_EvalFrameDefault
+ 19: _PyFunction_FastCall [('tile_f', [-1, 32, 1, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4109301
+ No: 14 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 8, 32]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,657998
+ No: 15 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 16, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,380668
+ No: 16 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 128, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10092522
+ No: 17 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 4, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2766793
+ No: 18 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 4, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7565072
+ No: 19 GFLOPS: 0.00/40.99 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
@@ -1766,9 +2512,8 @@ for this template
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
- tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 8, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1635073
- No: 19 GFLOPS: 16.89/161.19 result: MeasureResult(costs=(0.013710141181818182,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.34546685218811, timestamp=1682987936.1495934) [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5174693
- No: 20 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+ 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, 2, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2618857
+ No: 20 GFLOPS: 0.00/40.99 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
@@ -1890,7 +2635,7 @@ 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, 64, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1477150
+ tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 2, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,564976
@@ -1945,9 +2690,9 @@ and measure running time.
Finish loading 20 records
Best config:
- [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5295007
+ [('tile_f', [-1, 1, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2784230
Finish loading 20 records
- Time cost of this operator: 0.002261
+ Time cost of this operator: 0.006015
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 87539ea9f9..b5c955c5fe 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
@@ -360,10 +360,10 @@ 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 316.2 98.733 (1, 2, 10, 10, 3) 2 1 [316.2]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.099 0.968 (1, 6, 10, 10) 1 1 [3.099]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.959 0.299 (1, 1, 10, 10, 3) 1 1 [0.959]
- Total_time - 320.257 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 315.6 98.743 (1, 2, 10, 10, 3) 2 1 [315.6]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.052 0.955 (1, 6, 10, 10) 1 1 [3.052]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.966 0.302 (1, 1, 10, 10, 3) 1 1 [0.966]
+ Total_time - 319.618 - - - - -
@@ -428,10 +428,10 @@ 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 138.4 98.128 (1, 6, 10, 10, 1) 2 1 [138.4]
- tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.8 1.276 (1, 6, 10, 10) 1 1 [1.8]
- tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.841 0.596 (1, 3, 10, 10, 1) 1 1 [0.841]
- Total_time - 141.041 - - - - -
+ tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 101.3 97.383 (1, 6, 10, 10, 1) 2 1 [101.3]
+ tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.772 1.703 (1, 6, 10, 10) 1 1 [1.772]
+ tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.951 0.914 (1, 1, 10, 10, 3) 1 1 [0.951]
+ Total_time - 104.022 - - - - -
@@ -439,7 +439,7 @@ Timing the tuned program
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 25.585 seconds)
+ **Total running time of the script:** ( 1 minutes 24.486 seconds)
.. _sphx_glr_download_how_to_work_with_microtvm_micro_autotune.py:
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 7bd8b29cdf..19e92cec51 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
@@ -118,7 +118,7 @@ download a cat image and preprocess it to use as the model input.
/venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/ao/quantization/utils.py:281: UserWarning: must run observer before calling calculate_qparams. Returning default values.
"must run observer before calling calculate_qparams. " +
Downloading: "https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
-
0%| | 0.00/3.42M [00:00<?, ?B/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 72.9MB/s]
+
0%| | 0.00/3.42M [00:00<?, ?B/s]
62%|######2 | 2.13M/3.42M [00:00<00:00, 22.4MB/s]
100%|##########| 3.42M/3.42M [00:00<00:00, 33.8MB/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.
@@ -324,7 +324,7 @@ Look up prediction top 1 index in 1000 class synset.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 20.376 seconds)
+ **Total running time of the script:** ( 1 minutes 20.458 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 46af33572a..ea5eaf854f 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
@@ -217,7 +217,7 @@ take about **2 minutes** to download the Stanford Cars, while COCO 2017 validati
.. code-block:: none
- '/tmp/tmp293dho4g/images/random'
+ '/tmp/tmp0h6ga6xi/images/random'
@@ -308,7 +308,7 @@ 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], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0]
+ :alt: [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [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
@@ -317,8 +317,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
.. code-block:: none
- /tmp/tmp293dho4g/images/target contains 8144 images
- /tmp/tmp293dho4g/images/random contains 5000 images
+ /tmp/tmp0h6ga6xi/images/target contains 8144 images
+ /tmp/tmp0h6ga6xi/images/random contains 5000 images
@@ -493,13 +493,13 @@ the time on our validation set).
.. code-block:: none
Epoch 1/3
- 328/328 - 41s - loss: 0.2148 - accuracy: 0.9258 - val_loss: 0.2106 - val_accuracy: 0.9339 - 41s/epoch - 124ms/step
+ 328/328 - 40s - loss: 0.2085 - accuracy: 0.9287 - val_loss: 0.1437 - val_accuracy: 0.9562 - 40s/epoch - 122ms/step
Epoch 2/3
- 328/328 - 35s - loss: 0.0970 - accuracy: 0.9655 - val_loss: 0.1369 - val_accuracy: 0.9524 - 35s/epoch - 107ms/step
+ 328/328 - 35s - loss: 0.0874 - accuracy: 0.9671 - val_loss: 0.1198 - val_accuracy: 0.9539 - 35s/epoch - 108ms/step
Epoch 3/3
- 328/328 - 35s - loss: 0.0602 - accuracy: 0.9768 - val_loss: 0.1027 - val_accuracy: 0.9622 - 35s/epoch - 106ms/step
+ 328/328 - 35s - loss: 0.0651 - accuracy: 0.9783 - val_loss: 0.1155 - val_accuracy: 0.9641 - 35s/epoch - 107ms/step
- <keras.callbacks.History object at 0x7fceb538fbd0>
+ <keras.callbacks.History object at 0x7fa88e2ae510>
@@ -860,7 +860,7 @@ 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 32.894 seconds)
+ **Total running time of the script:** ( 4 minutes 28.517 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 57617c242f..90f858aa8b 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
@@ -5,20 +5,20 @@
Computation times
=================
-**07:45.319** total execution time for **how_to_work_with_microtvm** files:
+**07:39.833** total execution time for **how_to_work_with_microtvm** files:
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:32.894 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``) | 04:28.517 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 01:25.585 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``) | 01:24.486 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:20.376 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``) | 01:20.458 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:10.628 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``) | 00:10.629 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_custom_ide.py` (``micro_custom_ide.py``) | 00:08.325 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_custom_ide.py` (``micro_custom_ide.py``) | 00:08.270 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:07.512 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``) | 00:07.473 | 0.0 MB |
+-----------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_microtvm_micro_ethosu.py` (``micro_ethosu.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 40c773703a..bd227c6201 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
=================
-**00:37.268** total execution time for **how_to_work_with_relay** files:
+**00:37.443** total execution time for **how_to_work_with_relay** files:
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.528 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.468 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:02.931 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``) | 00:02.941 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:01.802 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``) | 00:02.028 | 0.0 MB |
+----------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``) | 00:00.007 | 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 ba7c6b9205..3982d8a2ca 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
@@ -278,7 +278,7 @@ The following example customizes CUDA lowering rule for :code:`exp`.
.. code-block:: none
- <function my_cuda_math_rule at 0x7fcafe5d4950>
+ <function my_cuda_math_rule at 0x7fa4d2daf200>
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 ecd1303f5f..7148080f59 100644
--- a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
@@ -5,22 +5,22 @@
Computation times
=================
-**00:09.334** total execution time for **how_to_work_with_schedules** files:
+**00:08.385** total execution time for **how_to_work_with_schedules** files:
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:06.570 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``) | 00:05.696 | 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_tensorize.py` (``tensorize.py``) | 00:01.190 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.624 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``) | 00:00.617 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.605 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``) | 00:00.603 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.136 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``) | 00:00.129 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.066 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``) | 00:00.064 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.058 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.056 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_how_to_work_with_schedules_tuple_inputs.py` (``tuple_inputs.py``) | 00:00.030 | 0.0 MB |
+------------------------------------------------------------------------------------------------+-----------+--------+
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 8f8cf70b49..98a084547b 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
=================
-**00:33.763** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:32.193** total execution time for **topic_vta_tutorials_autotvm** files:
+---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:33.756 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:32.186 | 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 59774f590c..77d5abc071 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 36.36s!
+ resnet18_v1 inference graph built in 34.31s!
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 97870e82e4..0a29accb7d 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 24.75s!
+ yolov3-tiny inference graph built in 23.56s!
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 4446a93f00..78480305e7 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
=================
-**01:45.377** total execution time for **topic_vta_tutorials_frontend** files:
+**01:42.082** total execution time for **topic_vta_tutorials_frontend** files:
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:53.331 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:51.172 | 0.0 MB |
+------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:52.046 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``) | 00:50.909 | 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 47f2c51f07..e1b9f5c3d0 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
=================
-**00:03.327** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.280** total execution time for **topic_vta_tutorials_optimize** files:
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.826 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``) | 00:02.797 | 0.0 MB |
+--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.500 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.483 | 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 eba04c4476..b3e2d6b31a 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
=================
-**00:00.868** total execution time for **topic_vta_tutorials** files:
+**00:00.814** total execution time for **topic_vta_tutorials** files:
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.450 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.420 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.418 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.394 | 0.0 MB |
+---------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index 0e4f711d4b..1d7cff72a3 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -318,7 +318,7 @@ We build the binary and check its correctness and performance.
.. code-block:: none
- Execution time of this operator: 95.255 ms
+ Execution time of this operator: 95.373 ms
@@ -434,7 +434,7 @@ operations.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 28.560 seconds)
+ **Total running time of the script:** ( 1 minutes 31.494 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 2641dfaba1..1d09e74e56 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: 11.02/11.02 result: MeasureResult(costs=(0.02434861,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7111151218414307, timestamp=1682986161.6233652) [('tile_y', [-1, 16]), ('tile_x', [-1, 256])],None,84
- No: 2 GFLOPS: 3.46/11.02 result: MeasureResult(costs=(0.077571358,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.5223381519317627, timestamp=1682986163.130107) [('tile_y', [-1, 4]), ('tile_x', [-1, 16])],None,42
- No: 3 GFLOPS: 3.76/11.02 result: MeasureResult(costs=(0.0713362332,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4039065837860107, timestamp=1682986165.8683152) [('tile_y', [-1, 16]), ('tile_x', [-1, 8])],None,34
- No: 4 GFLOPS: 3.27/11.02 result: MeasureResult(costs=(0.0822006338,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.5740251541137695, timestamp=1682986168.7850728) [('tile_y', [-1, 4]), ('tile_x', [-1, 8])],None,32
- No: 5 GFLOPS: 9.39/11.02 result: MeasureResult(costs=(0.0285809682,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7063653469085693, timestamp=1682986169.6943984) [('tile_y', [-1, 512]), ('tile_x', [-1, 32])],None,59
- No: 6 GFLOPS: 2.36/11.02 result: MeasureResult(costs=(0.1136443108,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.072852849960327, timestamp=1682986171.7710102) [('tile_y', [-1, 8]), ('tile_x', [-1, 2])],None,13
- No: 7 GFLOPS: 10.28/11.02 result: MeasureResult(costs=(0.026118831199999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7026386260986328, timestamp=1682986173.7827885) [('tile_y', [-1, 8]), ('tile_x', [-1, 32])],None,53
- No: 8 GFLOPS: 14.32/14.32 result: MeasureResult(costs=(0.018743647000000002,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6235649585723877, timestamp=1682986174.344845) [('tile_y', [-1, 32]), ('tile_x', [-1, 64])],None,65
- No: 9 GFLOPS: 3.48/14.32 result: MeasureResult(costs=(0.07708763360000001,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.454371452331543, timestamp=1682986175.9196217) [('tile_y', [-1, 2]), ('tile_x', [-1, 8])],None,31
- No: 10 GFLOPS: 3.61/14.32 result: MeasureResult(costs=(0.0744359562,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4187042713165283, timestamp=1682986177.3736377) [('tile_y', [-1, 128]), ('tile_x', [-1, 16])],None,47
+ No: 1 GFLOPS: 2.28/2.28 result: MeasureResult(costs=(0.11778225979999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.14568829536438, timestamp=1683000635.3602366) [('tile_y', [-1, 2]), ('tile_x', [-1, 8])],None,31
+ No: 2 GFLOPS: 2.16/2.28 result: MeasureResult(costs=(0.12420889199999999,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.2488672733306885, timestamp=1683000637.60532) [('tile_y', [-1, 128]), ('tile_x', [-1, 4])],None,27
+ No: 3 GFLOPS: 8.91/8.91 result: MeasureResult(costs=(0.0301308138,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.064992904663086, timestamp=1683000639.6331787) [('tile_y', [-1, 16]), ('tile_x', [-1, 128])],None,74
+ No: 4 GFLOPS: 13.19/13.19 result: MeasureResult(costs=(0.020347056199999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6390974521636963, timestamp=1683000641.508798) [('tile_y', [-1, 128]), ('tile_x', [-1, 128])],None,77
+ No: 5 GFLOPS: 11.87/13.19 result: MeasureResult(costs=(0.0226120654,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6879913806915283, timestamp=1683000642.315864) [('tile_y', [-1, 64]), ('tile_x', [-1, 32])],None,56
+ No: 6 GFLOPS: 11.62/13.19 result: MeasureResult(costs=(0.0231006776,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6305673122406006, timestamp=1683000642.9526894) [('tile_y', [-1, 128]), ('tile_x', [-1, 512])],None,97
+ No: 7 GFLOPS: 11.43/13.19 result: MeasureResult(costs=(0.0234894554,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6507046222686768, timestamp=1683000644.8711333) [('tile_y', [-1, 128]), ('tile_x', [-1, 256])],None,87
+ No: 8 GFLOPS: 4.54/13.19 result: MeasureResult(costs=(0.059063332,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2118299007415771, timestamp=1683000646.07284) [('tile_y', [-1, 16]), ('tile_x', [-1, 16])],None,44
+ No: 9 GFLOPS: 1.52/13.19 result: MeasureResult(costs=(0.1761398522,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.0669846534729004, timestamp=1683000649.2560663) [('tile_y', [-1, 64]), ('tile_x', [-1, 4])],None,26
+ No: 10 GFLOPS: 9.44/13.19 result: MeasureResult(costs=(0.028431742200000006,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6832067966461182, timestamp=1683000649.975619) [('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 502bcf22f2..0a21d4bc98 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': 517.9155723499991, 'median': 518.3068678500035, 'std': 1.5242307408517035}
+ {'mean': 512.0114532300022, 'median': 512.5405834499986, 'std': 2.839063755927517}
@@ -582,31 +582,31 @@ 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: 13.40/ 22.46 GFLOPS | Progress: (4/20) | 12.58 s
[Task 1/25] Current/Best: 13.52/ 22.46 GFLOPS | Progress: (8/20) | 16.20 s
[Task 1/25] Current/Best: 11.27/ 22.56 GFLOPS | Progress: (12/20) | 20.22 s
[Task 1/25] Current/Best: 1.91/ 22.56 GFLOPS | Progress: (16/20) | 24.39 s
[Task 1/25] Current/Best: 15.03/ 23.25 GFLOPS | Progress: (20/20) | 26.53 s Done.
-
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 16.90/ 16.90 GFLOPS | Progress: (4/20) | 5.10 s
[Task 2/25] Current/Best: 13.77/ 17.23 GFLOPS | Progress: (8/20) | 7.24 s
[Task 2/25] Current/Best: 12.95/ 18.13 GFLOPS | Progress: (12/20) | 8.81 s
[Task 2/25] Current/Best: 12.92/ 18.13 GFLOPS | Progress: (16/20) | 10.59 s
[Task 2/25] Current/Best: 6.62/ 18.13 GFLOPS | Progress: (20/20) | 12.00 s Done.
-
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 9.99/ 19.76 GFLOPS | Progress: (4/20) | 5.60 s
[Task 3/25] Current/Best: 17.48/ 19.76 GFLOPS | Progress: (8/20) | 8.73 s
[Task 3/25] Current/Best: 17.63/ 22.76 GFLOPS | Progress: (12/20) | 10.61 s
[Task 3/25] Current/Best: 12.45/ 22.76 GFLOPS | Progress: (16/20) | 13.63 s
[Task 3/25] Current/Best: 8.68/ 22.76 GFLOPS | Progress: (20/20) | 16.24 s Done.
-
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 9.64/ 9.65 GFLOPS | Progress: (4/20) | 7.91 s
[Task 4/25] Current/Best: 16.35/ 16.35 GFLOPS | Progress: (8/20) | 10.54 s
[Task 4/25] Current/Best: 20.95/ 20.95 GFLOPS | Progress: (12/20) | 12.61 s
[Task 4/25] Current/Best: 6.26/ 20.95 GFLOPS | Progress: (16/20) | 16.14 s
[Task 4/25] Current/Best: 3.78/ 20.95 GFLOPS | Progress: (20/20) | 19.41 s Done.
-
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 10.35/ 17.35 GFLOPS | Progress: (4/20) | 5.10 s
[Task 5/25] Current/Best: 13.33/ 17.54 GFLOPS | Progress: (8/20) | 7.86 s
[Task 5/25] Current/Best: 17.53/ 18.12 GFLOPS | Progress: (12/20) | 9.89 s
[Task 5/25] Current/Best: 22.91/ 22.91 GFLOPS | Progress: (16/20) | 12.06 s
[Task 5/25] Current/Best: 3.06/ 22.91 GFLOPS | Progress: (20/20) | 14.41 s Done.
-
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 6.79/ 13.68 GFLOPS | Progress: (4/20) | 7.16 s
[Task 6/25] Current/Best: 18.50/ 19.51 GFLOPS | Progress: (8/20) | 10.41 s
[Task 6/25] Current/Best: 9.90/ 19.51 GFLOPS | Progress: (12/20) | 14.25 s
[Task 6/25] Current/Best: 14.04/ 19.51 GFLOPS | Progress: (16/20) | 17.52 s
[Task 6/25] Current/Best: 6.52/ 19.51 GFLOPS | Progress: (20/20) | 21.08 s Done.
-
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 17.39/ 17.39 GFLOPS | Progress: (4/20) | 6.84 s
[Task 7/25] Current/Best: 7.75/ 17.39 GFLOPS | Progress: (8/20) | 9.50 s
[Task 7/25] Current/Best: 17.04/ 17.39 GFLOPS | Progress: (12/20) | 12.19 s
[Task 7/25] Current/Best: 15.73/ 17.39 GFLOPS | Progress: (16/20) | 14.82 s
[Task 7/25] Current/Best: 16.28/ 18.96 GFLOPS | Progress: (20/20) | 17.12 s Done.
-
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 2.92/ 16.07 GFLOPS | Progress: (4/20) | 6.66 s
[Task 8/25] Current/Best: 15.88/ 16.07 GFLOPS | Progress: (8/20) | 9.96 s
[Task 8/25] Current/Best: 9.98/ 16.07 GFLOPS | Progress: (12/20) | 19.82 s
[Task 8/25] Current/Best: 10.70/ 16.07 GFLOPS | Progress: (16/20) | 22.55 s
[Task 8/25] Current/Best: 12.85/ 17.11 GFLOPS | Progress: (20/20) | 24.91 s Done.
-
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 16.53/ 18.09 GFLOPS | Progress: (4/20) | 4.67 s
[Task 9/25] Current/Best: 8.58/ 18.09 GFLOPS | Progress: (8/20) | 12.15 s
[Task 9/25] Current/Best: 18.33/ 18.33 GFLOPS | Progress: (12/20) | 14.85 s
[Task 9/25] Current/Best: 14.73/ 18.33 GFLOPS | Progress: (16/20) | 17.21 s
[Task 9/25] Current/Best: 2.72/ 20.70 GFLOPS | Progress: (20/20) | 19.16 s Done.
-
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 17.63/ 17.63 GFLOPS | Progress: (4/20) | 7.10 s
[Task 10/25] Current/Best: 3.82/ 17.63 GFLOPS | Progress: (8/20) | 9.12 s
[Task 10/25] Current/Best: 5.38/ 17.63 GFLOPS | Progress: (12/20) | 11.74 s
[Task 10/25] Current/Best: 4.78/ 17.63 GFLOPS | Progress: (16/20) | 14.09 s
[Task 10/25] Current/Best: 9.97/ 17.63 GFLOPS | Progress: (20/20) | 16.52 s Done.
-
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 11.16/ 23.03 GFLOPS | Progress: (4/20) | 5.42 s
[Task 11/25] Current/Best: 17.22/ 23.03 GFLOPS | Progress: (8/20) | 7.37 s
[Task 11/25] Current/Best: 11.64/ 23.03 GFLOPS | Progress: (12/20) | 10.01 s
[Task 11/25] Current/Best: 12.25/ 23.03 GFLOPS | Progress: (16/20) | 12.42 s
[Task 11/25] Current/Best: 18.09/ 23.03 GFLOPS | Progress: (20/20) | 15.33 s Done.
-
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 11.12/ 18.23 GFLOPS | Progress: (4/20) | 6.54 s
[Task 12/25] Current/Best: 2.61/ 18.23 GFLOPS | Progress: (8/20) | 9.52 s
[Task 12/25] Current/Best: 14.81/ 18.79 GFLOPS | Progress: (12/20) | 12.47 s
[Task 12/25] Current/Best: 10.62/ 18.79 GFLOPS | Progress: (16/20) | 17.67 s
[Task 12/25] Current/Best: 21.11/ 21.11 GFLOPS | Progress: (20/20) | 19.84 s Done.
-
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 17.04/ 19.23 GFLOPS | Progress: (4/20) | 5.08 s
[Task 13/25] Current/Best: 11.88/ 22.37 GFLOPS | Progress: (8/20) | 8.82 s
[Task 13/25] Current/Best: 15.17/ 22.37 GFLOPS | Progress: (12/20) | 11.72 s
[Task 13/25] Current/Best: 18.59/ 22.37 GFLOPS | Progress: (16/20) | 14.37 s
[Task 13/25] Current/Best: 18.04/ 22.37 GFLOPS | Progress: (20/20) | 16.83 s Done.
-
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 10.46/ 17.72 GFLOPS | Progress: (4/20) | 6.92 s
[Task 14/25] Current/Best: 22.25/ 22.25 GFLOPS | Progress: (8/20) | 9.13 s
[Task 14/25] Current/Best: 17.00/ 22.25 GFLOPS | Progress: (12/20) | 12.18 s
[Task 14/25] Current/Best: 17.40/ 22.25 GFLOPS | Progress: (16/20) | 17.03 s
[Task 14/25] Current/Best: 13.47/ 22.25 GFLOPS | Progress: (20/20) | 19.57 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 15/25] Current/Best: 9.10/ 17.61 GFLOPS | Progress: (4/20) | 10.72 s
[Task 15/25] Current/Best: 16.15/ 17.61 GFLOPS | Progress: (8/20) | 13.47 s Done.
-
[Task 15/25] Current/Best: 18.87/ 18.87 GFLOPS | Progress: (12/20) | 18.33 s
[Task 15/25] Current/Best: 8.58/ 18.87 GFLOPS | Progress: (16/20) | 25.12 s
[Task 15/25] Current/Best: 15.85/ 18.87 GFLOPS | Progress: (20/20) | 27.10 s Done.
-
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 9.89/ 17.94 GFLOPS | Progress: (4/20) | 4.65 s
[Task 16/25] Current/Best: 18.22/ 18.22 GFLOPS | Progress: (8/20) | 8.46 s
[Task 16/25] Current/Best: 7.44/ 18.22 GFLOPS | Progress: (12/20) | 11.62 s
[Task 16/25] Current/Best: 14.67/ 18.54 GFLOPS | Progress: (16/20) | 13.51 s
[Task 16/25] Current/Best: 18.46/ 18.54 GFLOPS | Progress: (20/20) | 15.51 s Done.
-
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 12.17/ 17.77 GFLOPS | Progress: (4/20) | 5.66 s
[Task 17/25] Current/Best: 19.19/ 19.19 GFLOPS | Progress: (8/20) | 8.07 s
[Task 17/25] Current/Best: 3.08/ 19.19 GFLOPS | Progress: (12/20) | 11.57 s
[Task 17/25] Current/Best: 23.62/ 23.62 GFLOPS | Progress: (16/20) | 14.23 s
[Task 17/25] Current/Best: 15.05/ 23.62 GFLOPS | Progress: (20/20) | 16.89 s Done.
-
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 5.61/ 13.07 GFLOPS | Progress: (4/20) | 8.24 s
[Task 18/25] Current/Best: 10.14/ 20.31 GFLOPS | Progress: (8/20) | 11.02 s
[Task 18/25] Current/Best: 14.07/ 20.57 GFLOPS | Progress: (12/20) | 13.50 s
[Task 18/25] Current/Best: 14.99/ 20.57 GFLOPS | Progress: (16/20) | 15.41 s
[Task 18/25] Current/Best: 9.42/ 20.57 GFLOPS | Progress: (20/20) | 17.91 s Done.
-
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 20.01/ 20.01 GFLOPS | Progress: (4/20) | 6.35 s
[Task 19/25] Current/Best: 12.16/ 20.01 GFLOPS | Progress: (8/20) | 10.60 s
[Task 19/25] Current/Best: 22.43/ 22.43 GFLOPS | Progress: (12/20) | 14.13 s
[Task 19/25] Current/Best: 21.70/ 22.43 GFLOPS | Progress: (16/20) | 18.00 s
[Task 19/25] Current/Best: 17.33/ 22.43 GFLOPS | Progress: (20/20) | 20.32 s Done.
-
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 5.20/ 19.03 GFLOPS | Progress: (4/20) | 5.83 s
[Task 20/25] Current/Best: 8.35/ 19.03 GFLOPS | Progress: (8/20) | 7.73 s
[Task 20/25] Current/Best: 13.76/ 19.03 GFLOPS | Progress: (12/20) | 12.33 s
[Task 20/25] Current/Best: 16.24/ 19.03 GFLOPS | Progress: (16/20) | 15.40 s
[Task 20/25] Current/Best: 4.59/ 19.15 GFLOPS | Progress: (20/20) | 17.79 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 22.29/ 22.29 GFLOPS | Progress: (4/20) | 5.91 s
[Task 21/25] Current/Best: 10.49/ 22.29 GFLOPS | Progress: (8/20) | 7.72 s
[Task 21/25] Current/Best: 16.28/ 22.29 GFLOPS | Progress: (12/20) | 10.49 s
[Task 21/25] Current/Best: 17.81/ 22.29 GFLOPS | Progress: (16/20) | 14.00 s
[Task 21/25] Current/Best: 14.14/ 22.29 GFLOPS | Progress: (20/20
) | 15.85 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
[Task 1/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 1/25] Current/Best: 7.75/ 17.69 GFLOPS | Progress: (4/20) | 10.32 s
[Task 1/25] Current/Best: 15.04/ 22.75 GFLOPS | Progress: (8/20) | 15.51 s
[Task 1/25] Current/Best: 6.26/ 23.40 GFLOPS | Progress: (12/20) | 17.94 s
[Task 1/25] Current/Best: 13.62/ 23.40 GFLOPS | Progress: (16/20) | 20.79 s
[Task 1/25] Current/Best: 12.45/ 23.40 GFLOPS | Progress: (20/20) | 25.34 s Done.
+
[Task 2/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 2/25] Current/Best: 12.68/ 16.88 GFLOPS | Progress: (4/20) | 4.81 s
[Task 2/25] Current/Best: 8.83/ 21.11 GFLOPS | Progress: (8/20) | 6.49 s
[Task 2/25] Current/Best: 13.05/ 21.11 GFLOPS | Progress: (12/20) | 8.53 s
[Task 2/25] Current/Best: 17.42/ 21.11 GFLOPS | Progress: (16/20) | 9.81 s
[Task 2/25] Current/Best: 8.43/ 21.11 GFLOPS | Progress: (20/20) | 11.45 s Done.
+
[Task 3/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 3/25] Current/Best: 9.88/ 15.31 GFLOPS | Progress: (4/20) | 6.72 s
[Task 3/25] Current/Best: 6.82/ 21.27 GFLOPS | Progress: (8/20) | 8.88 s
[Task 3/25] Current/Best: 10.96/ 21.72 GFLOPS | Progress: (12/20) | 12.61 s
[Task 3/25] Current/Best: 8.16/ 21.72 GFLOPS | Progress: (16/20) | 15.26 s
[Task 3/25] Current/Best: 20.34/ 21.72 GFLOPS | Progress: (20/20) | 18.78 s Done.
+
[Task 4/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 4/25] Current/Best: 13.12/ 20.75 GFLOPS | Progress: (4/20) | 4.64 s
[Task 4/25] Current/Best: 9.76/ 20.75 GFLOPS | Progress: (8/20) | 7.62 s
[Task 4/25] Current/Best: 6.29/ 20.75 GFLOPS | Progress: (12/20) | 10.58 s
[Task 4/25] Current/Best: 19.13/ 20.75 GFLOPS | Progress: (16/20) | 15.84 s
[Task 4/25] Current/Best: 17.57/ 20.75 GFLOPS | Progress: (20/20) | 18.19 s Done.
+
[Task 5/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 5/25] Current/Best: 11.88/ 14.77 GFLOPS | Progress: (4/20) | 5.53 s
[Task 5/25] Current/Best: 11.44/ 14.77 GFLOPS | Progress: (8/20) | 7.76 s
[Task 5/25] Current/Best: 13.57/ 17.48 GFLOPS | Progress: (12/20) | 9.69 s
[Task 5/25] Current/Best: 14.22/ 18.19 GFLOPS | Progress: (16/20) | 11.56 s
[Task 5/25] Current/Best: 9.50/ 20.45 GFLOPS | Progress: (20/20) | 13.80 s Done.
+
[Task 6/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 6/25] Current/Best: 7.29/ 15.90 GFLOPS | Progress: (4/20) | 5.40 s
[Task 6/25] Current/Best: 11.29/ 15.90 GFLOPS | Progress: (8/20) | 11.00 s
[Task 6/25] Current/Best: 16.51/ 16.51 GFLOPS | Progress: (12/20) | 13.72 s
[Task 6/25] Current/Best: 11.49/ 16.51 GFLOPS | Progress: (16/20) | 16.22 s
[Task 6/25] Current/Best: 17.80/ 17.80 GFLOPS | Progress: (20/20) | 20.03 s Done.
+
[Task 7/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 7/25] Current/Best: 12.96/ 12.96 GFLOPS | Progress: (4/20) | 5.53 s
[Task 7/25] Current/Best: 6.78/ 18.89 GFLOPS | Progress: (8/20) | 8.36 s
[Task 7/25] Current/Best: 16.73/ 18.89 GFLOPS | Progress: (12/20) | 11.09 s
[Task 7/25] Current/Best: 15.01/ 18.89 GFLOPS | Progress: (16/20) | 13.53 s
[Task 7/25] Current/Best: 11.21/ 18.89 GFLOPS | Progress: (20/20) | 16.86 s Done.
+
[Task 8/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 8/25] Current/Best: 8.48/ 15.03 GFLOPS | Progress: (4/20) | 5.32 s
[Task 8/25] Current/Best: 2.49/ 15.03 GFLOPS | Progress: (8/20) | 10.26 s
[Task 8/25] Current/Best: 10.76/ 15.03 GFLOPS | Progress: (12/20) | 21.97 s
[Task 8/25] Current/Best: 10.18/ 15.03 GFLOPS | Progress: (16/20) | 26.45 s
[Task 8/25] Current/Best: 11.60/ 15.03 GFLOPS | Progress: (20/20) | 36.21 s
[Task 9/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 9/25] Current/Best: 8.19/ 14.58 GFLOPS | Progress: (4/20) | 6.14 s
[Task 9/25] Current/Best: 9.41/ 21.93 GFLOPS | Progress: (8/20) | 10.89 s
[Task 9/25] Current/Best: 18.56/ 21.93 GFLOPS | Progress: (12/20) | 15.22 s
[Task 9/25] Current/Best: 16.10/ 21.93 GFLOPS | Progress: (16/20) | 20.48 s
[Task 9/25] Current/Best: 7.35/ 21.93 GFLOPS | Progress: (20/
20) | 22.93 s Done.
+
[Task 10/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 10/25] Current/Best: 15.37/ 20.92 GFLOPS | Progress: (4/20) | 5.21 s
[Task 10/25] Current/Best: 6.36/ 20.92 GFLOPS | Progress: (8/20) | 7.32 s
[Task 10/25] Current/Best: 20.81/ 20.92 GFLOPS | Progress: (12/20) | 10.00 s
[Task 10/25] Current/Best: 2.94/ 20.92 GFLOPS | Progress: (16/20) | 12.49 s
[Task 10/25] Current/Best: 16.30/ 20.92 GFLOPS | Progress: (20/20) | 14.53 s Done.
+
[Task 11/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 11/25] Current/Best: 11.41/ 14.10 GFLOPS | Progress: (4/20) | 5.42 s
[Task 11/25] Current/Best: 16.89/ 16.89 GFLOPS | Progress: (8/20) | 9.30 s
[Task 11/25] Current/Best: 9.03/ 16.89 GFLOPS | Progress: (12/20) | 12.65 s
[Task 11/25] Current/Best: 16.56/ 16.89 GFLOPS | Progress: (16/20) | 14.95 s
[Task 11/25] Current/Best: 9.81/ 23.73 GFLOPS | Progress: (20/20) | 18.13 s Done.
+
[Task 12/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 12/25] Current/Best: 14.30/ 19.65 GFLOPS | Progress: (4/20) | 7.78 s
[Task 12/25] Current/Best: 18.12/ 19.65 GFLOPS | Progress: (8/20) | 9.95 s
[Task 12/25] Current/Best: 14.37/ 19.65 GFLOPS | Progress: (12/20) | 13.02 s
[Task 12/25] Current/Best: 12.82/ 19.65 GFLOPS | Progress: (16/20) | 15.45 s
[Task 12/25] Current/Best: 14.51/ 19.65 GFLOPS | Progress: (20/20) | 18.02 s Done.
+
[Task 13/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 13/25] Current/Best: 19.35/ 20.10 GFLOPS | Progress: (4/20) | 5.26 s
[Task 13/25] Current/Best: 18.13/ 20.10 GFLOPS | Progress: (8/20) | 9.02 s
[Task 13/25] Current/Best: 9.31/ 20.10 GFLOPS | Progress: (12/20) | 12.37 s
[Task 13/25] Current/Best: 9.14/ 20.10 GFLOPS | Progress: (16/20) | 14.80 s
[Task 13/25] Current/Best: 16.65/ 20.10 GFLOPS | Progress: (20/20) | 18.10 s Done.
+
[Task 14/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 14/25] Current/Best: 14.54/ 15.70 GFLOPS | Progress: (4/20) | 5.03 s
[Task 14/25] Current/Best: 8.67/ 16.25 GFLOPS | Progress: (8/20) | 8.08 s
[Task 14/25] Current/Best: 6.38/ 16.25 GFLOPS | Progress: (12/20) | 11.92 s
[Task 14/25] Current/Best: 13.53/ 16.25 GFLOPS | Progress: (16/20) | 14.99 s
[Task 14/25] Current/Best: 9.54/ 16.25 GFLOPS | Progress: (20/20) | 18.18 s
[Task 15/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
Done.
-
[Task 22/25] Current/Best: 17.77/ 17.88 GFLOPS | Progress: (4/20) | 6.85 s
[Task 22/25] Current/Best: 10.94/ 21.18 GFLOPS | Progress: (8/20) | 10.39 s
[Task 22/25] Current/Best: 11.94/ 21.18 GFLOPS | Progress: (12/20) | 12.23 s
[Task 22/25] Current/Best: 10.74/ 21.18 GFLOPS | Progress: (16/20) | 14.64 s
[Task 22/25] Current/Best: 10.37/ 21.18 GFLOPS | Progress: (20/20) | 17.43 s Done.
-
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 22.23/ 22.23 GFLOPS | Progress: (4/20) | 6.35 s
[Task 23/25] Current/Best: 4.99/ 22.23 GFLOPS | Progress: (8/20) | 10.53 s
[Task 23/25] Current/Best: 10.15/ 22.23 GFLOPS | Progress: (12/20) | 13.48 s
[Task 23/25] Current/Best: 18.99/ 22.23 GFLOPS | Progress: (16/20) | 18.68 s
[Task 23/25] Current/Best: 21.44/ 22.23 GFLOPS | Progress: (20/20) | 20.97 s Done.
-
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 0.91/ 2.88 GFLOPS | Progress: (4/20) | 13.70 s
[Task 24/25] Current/Best: 3.22/ 3.22 GFLOPS | Progress: (8/20) | 26.82 s
[Task 24/25] Current/Best: 1.69/ 4.08 GFLOPS | Progress: (12/20) | 37.85 s
[Task 24/25] Current/Best: 3.85/ 5.90 GFLOPS | Progress: (16/20) | 48.82 s
[Task 24/25] Current/Best: 6.56/ 9.69 GFLOPS | Progress: (20/20) | 60.86 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
[Task 25/25] Current/Best: 4.55/ 8.84 GFLOPS | Progress: (4/20) | 7.47 s
[Task 25/25] Current/Best: 5.74/ 8.84 GFLOPS | Progress: (8/20) | 18.48 s
[Task 25/25] Current/Best: 7.30/ 9.28 GFLOPS | Progress: (12/20) | 21.34 s
[Task 25/25] Current/Best: 1.54/ 9.28 GFLOPS | Progress: (16/20) | 24.98 s
[Task 25/25] Current/Best: 8.97/ 9.32 GFLOPS | Progress: (20/20) | 35.97 s
+
[Task 15/25] Current/Best: 18.75/ 18.75 GFLOPS | Progress: (4/20) | 6.03 s
[Task 15/25] Current/Best: 19.47/ 19.47 GFLOPS | Progress: (8/20) | 8.16 s
[Task 15/25] Current/Best: 16.30/ 20.14 GFLOPS | Progress: (12/20) | 9.66 s
[Task 15/25] Current/Best: 1.72/ 20.14 GFLOPS | Progress: (16/20) | 13.17 s
[Task 15/25] Current/Best: 20.57/ 20.57 GFLOPS | Progress: (20/20) | 15.08 s
[Task 16/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 16/25] Current/Best: 9.72/ 11.12 GFLOPS | Progress: (4/20) | 5.86 s
[Task 16/25] Current/Best: 6.32/ 13.29 GFLOPS | Progress: (8/20) | 8.20 s
[Task 16/25] Current/Best: 10.10/ 15.52 GFLOPS | Progress: (12/20) | 9.94 s
[Task 16/25] Current/Best: 10.07/ 17.86 GFLOPS | Progress: (16/20) | 12.61 s
[Task 16/25] Current/Best: 4.26/ 17.86 GFLOPS | Progress: (20/20) | 15.91 s Done.
+
[Task 17/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 17/25] Current/Best: 1.56/ 12.25 GFLOPS | Progress: (4/20) | 8.03 s
[Task 17/25] Current/Best: 14.85/ 22.52 GFLOPS | Progress: (8/20) | 10.67 s
[Task 17/25] Current/Best: 17.90/ 22.52 GFLOPS | Progress: (12/20) | 13.05 s
[Task 17/25] Current/Best: 23.57/ 23.57 GFLOPS | Progress: (16/20) | 15.17 s
[Task 17/25] Current/Best: 13.89/ 23.57 GFLOPS | Progress: (20/20) | 17.59 s Done.
+
[Task 18/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 18/25] Current/Best: 15.03/ 15.03 GFLOPS | Progress: (4/20) | 6.81 s
[Task 18/25] Current/Best: 13.17/ 19.94 GFLOPS | Progress: (8/20) | 8.89 s
[Task 18/25] Current/Best: 10.02/ 19.94 GFLOPS | Progress: (12/20) | 14.58 s
[Task 18/25] Current/Best: 13.90/ 19.94 GFLOPS | Progress: (16/20) | 18.75 s
[Task 18/25] Current/Best: 18.63/ 19.94 GFLOPS | Progress: (20/20) | 24.09 s Done.
+
[Task 19/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 19/25] Current/Best: 2.67/ 18.98 GFLOPS | Progress: (4/20) | 6.08 s
[Task 19/25] Current/Best: 12.01/ 18.98 GFLOPS | Progress: (8/20) | 9.01 s
[Task 19/25] Current/Best: 11.00/ 18.98 GFLOPS | Progress: (12/20) | 13.17 s
[Task 19/25] Current/Best: 11.54/ 18.98 GFLOPS | Progress: (16/20) | 16.02 s
[Task 19/25] Current/Best: 5.35/ 18.98 GFLOPS | Progress: (20/20) | 19.30 s Done.
+
[Task 20/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 20/25] Current/Best: 16.96/ 16.96 GFLOPS | Progress: (4/20) | 6.37 s
[Task 20/25] Current/Best: 11.62/ 16.96 GFLOPS | Progress: (8/20) | 10.14 s
[Task 20/25] Current/Best: 7.02/ 16.96 GFLOPS | Progress: (12/20) | 14.20 s
[Task 20/25] Current/Best: 10.46/ 16.96 GFLOPS | Progress: (16/20) | 16.60 s
[Task 20/25] Current/Best: 16.38/ 17.14 GFLOPS | Progress: (20/20) | 18.70 s
[Task 21/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 21/25] Current/Best: 22.14/ 22.14 GFLOPS | Progress: (4/20) | 4.40 s
[Task 21/25] Current/Best: 14.61/ 22.14 GFLOPS | Progress: (8/20) | 7.83 s
[Task 21/25] Current/Best: 11.63/ 22.14 GFLOPS | Progress: (12/20) | 10.13 s Done.
+ Done.
+
[Task 21/25] Current/Best: 14.08/ 22.14 GFLOPS | Progress: (16/20) | 12.98 s
[Task 21/25] Current/Best: 11.72/ 22.14 GFLOPS | Progress: (20/20) | 15.08 s
[Task 22/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 22/25] Current/Best: 5.06/ 15.79 GFLOPS | Progress: (4/20) | 5.90 s
[Task 22/25] Current/Best: 20.59/ 20.59 GFLOPS | Progress: (8/20) | 9.08 s
[Task 22/25] Current/Best: 19.59/ 20.59 GFLOPS | Progress: (12/20) | 10.83 s
[Task 22/25] Current/Best: 20.24/ 20.59 GFLOPS | Progress: (16/20) | 13.26 s
[Task 22/25] Current/Best: 4.46/ 20.59 GFLOPS | Progress: (20/20) | 15.15 s Done.
+
[Task 23/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 23/25] Current/Best: 20.10/ 20.10 GFLOPS | Progress: (4/20) | 9.11 s
[Task 23/25] Current/Best: 10.42/ 20.21 GFLOPS | Progress: (8/20) | 12.26 s
[Task 23/25] Current/Best: 6.14/ 20.21 GFLOPS | Progress: (12/20) | 16.19 s
[Task 23/25] Current/Best: 20.33/ 22.12 GFLOPS | Progress: (16/20) | 19.13 s
[Task 23/25] Current/Best: 10.88/ 22.12 GFLOPS | Progress: (20/20) | 23.02 s Done.
+
[Task 24/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s
[Task 24/25] Current/Best: 7.66/ 7.66 GFLOPS | Progress: (4/20) | 13.80 s
[Task 24/25] Current/Best: 3.25/ 7.66 GFLOPS | Progress: (8/20) | 24.78 s
[Task 24/25] Current/Best: 2.31/ 10.22 GFLOPS | Progress: (12/20) | 32.13 s
[Task 24/25] Current/Best: 3.16/ 10.22 GFLOPS | Progress: (16/20) | 43.07 s
[Task 24/25] Current/Best: 6.43/ 10.22 GFLOPS | Progress: (20/20) | 53.75 s
[Task 25/25] Current/Best: 0.00/ 0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+ Done.
+
[Task 25/25] Current/Best: 7.73/ 8.22 GFLOPS | Progress: (4/20) | 13.52 s
[Task 25/25] Current/Best: 3.01/ 8.69 GFLOPS | Progress: (8/20) | 25.89 s
[Task 25/25] Current/Best: 7.84/ 8.69 GFLOPS | Progress: (12/20) | 38.18 s
[Task 25/25] Current/Best: 6.91/ 8.69 GFLOPS | Progress: (16/20) | 50.16 s
[Task 25/25] Current/Best: 7.51/ 8.69 GFLOPS | Progress: (20/20) | 56.72 s
@@ -702,8 +702,8 @@ 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
@@ -760,8 +760,8 @@ improvement in comparing the optimized model to the unoptimized model.
.. code-block:: none
- optimized: {'mean': 412.96504630999834, 'median': 410.0456652999924, 'std': 4.772991813347499}
- unoptimized: {'mean': 517.9155723499991, 'median': 518.3068678500035, 'std': 1.5242307408517035}
+ optimized: {'mean': 417.7781961399978, 'median': 417.4723211000014, 'std': 1.603164789793621}
+ unoptimized: {'mean': 512.0114532300022, 'median': 512.5405834499986, 'std': 2.839063755927517}
@@ -784,7 +784,7 @@ profiling/benchmarking.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 12 minutes 46.823 seconds)
+ **Total running time of the script:** ( 13 minutes 0.393 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 09d68d8e06..a12bdc479b 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
- 1.247e-07 secs/op
+ 1.355e-07 secs/op
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index f98c6feed9..2b411d32dd 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -270,7 +270,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
.. code-block:: none
- [stage(a, placeholder(a, 0x72bd7b0)), stage(b, placeholder(b, 0x12ce20e0)), 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, 0x10bd2ca0)), stage(b, placeholder(b, 0x66366e0)), 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. [...]
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index eefb922255..0b1d4259a8 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,24 +5,24 @@
Computation times
=================
-**16:21.111** total execution time for **tutorial** files:
+**16:36.868** total execution time for **tutorial** files:
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 12:46.823 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``) | 13:00.393 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:28.560 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:31.494 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:00.838 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``) | 01:00.654 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:38.684 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``) | 00:37.774 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:23.352 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``) | 00:23.774 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.758 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``) | 00:01.724 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.896 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``) | 00:00.876 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.200 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.178 | 0.0 MB |
+------------------------------------------------------------------------------------------+-----------+--------+
| :ref:`sphx_glr_tutorial_uma.py` (``uma.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 ffa66190b8..04d3cfdf11 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -285,8 +285,8 @@ helper function to run a profile of the TVM generated code.
.. code-block:: none
- Numpy running time: 0.000008
- naive: 0.000007
+ Numpy running time: 0.000007
+ naive: 0.000008
@@ -444,7 +444,7 @@ factor to be the number of threads on your CPU.
.. code-block:: none
- vector: 0.000042
+ vector: 0.000040
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -498,10 +498,10 @@ We can now compare the different schedules
.. code-block:: none
Operator Timing Performance
- numpy 8.099830001810914e-06 1.0
- naive 6.6542e-06 0.8215234145052787
- parallel 6.9454000000000005e-06 0.8574747863161557
- vector 4.23105e-05 5.223628149052567
+ numpy 7.148350000534265e-06 1.0
+ naive 7.801199999999999e-06 1.091328768095706
+ parallel 7.021600000000001e-06 0.9822686353459482
+ vector 4.03934e-05 5.6507305877553575
@@ -922,7 +922,7 @@ matrix multiplication.
.. code-block:: none
- Numpy running time: 0.019415
+ Numpy running time: 0.018171
@@ -980,7 +980,7 @@ optimizations.
.. code-block:: none
- none: 3.327564
+ none: 3.389477
@@ -1080,7 +1080,7 @@ schedule.
.. code-block:: none
- blocking: 0.334628
+ blocking: 0.299349
@@ -1164,7 +1164,7 @@ already cache friendly from our previous optimizations.
.. code-block:: none
- vectorization: 0.357461
+ vectorization: 0.335625
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1230,7 +1230,7 @@ more cache friendly.
.. code-block:: none
- loop permutation: 0.120386
+ loop permutation: 0.116772
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1321,7 +1321,7 @@ optimized schedule.
.. code-block:: none
- array packing: 0.109156
+ array packing: 0.108581
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1404,7 +1404,7 @@ to `C` when all the block results are ready.
.. code-block:: none
- block caching: 0.111014
+ block caching: 0.110944
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1478,7 +1478,7 @@ of thread-level parallelization.
.. code-block:: none
- parallelization: 0.147060
+ parallelization: 0.146414
# from tvm.script import ir as I
# from tvm.script import tir as T
@@ -1548,13 +1548,13 @@ working, we can compare the results.
.. code-block:: none
Operator Timing Performance
- none 3.3275638661999998 1.0
- blocking 0.33462796669999995 0.10056244753076288
- vectorization 0.3574610829 0.10742425908964212
- loop permutation 0.1203855278 0.036178277154294704
- array packing 0.1091556324 0.03280346727789576
- block caching 0.11101354869999999 0.03336180856741147
- parallelization 0.14706005230000002 0.04419450932069988
+ none 3.3894773024000004 1.0
+ blocking 0.29934911880000004 0.0883171923258016
+ vectorization 0.3356248882 0.09901965945084004
+ loop permutation 0.1167720905 0.03445135638386389
+ array packing 0.10858137099999998 0.03203484234076928
+ block caching 0.1109444232 0.032732015382266506
+ parallelization 0.14641371099999997 0.043196545643284956
@@ -1596,7 +1596,7 @@ the computation for specific platforms.
.. rst-class:: sphx-glr-timing
- **Total running time of the script:** ( 1 minutes 0.838 seconds)
+ **Total running time of the script:** ( 1 minutes 0.654 seconds)
.. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
diff --git a/docs/commit_hash b/docs/commit_hash
index 7891f60600..5320fd247b 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-9530a8de1e2e6d150748ee6a7b45d39195785f50
+c1e88b365a240a99452b33f0fe30776e01629125
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index 021a3c0eb6..be972687bf 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -590,7 +590,7 @@ 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 27.616 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 24.849 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_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index 2e9a2d2057..148cda8974 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -444,7 +444,7 @@
<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.zip995aba3a-7257-4a91-8e80-24528214d098 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.zipf60f7196-0da9-4273-afb2-2b6a6e4c8534 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 c112b5acb5..4987892f4b 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -454,14 +454,13 @@ 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]
- 19%|#9 | 8.00M/41.5M [00:00<00:01, 31.5MB/s]
- 39%|###8 | 16.0M/41.5M [00:00<00:00, 35.7MB/s]
- 54%|#####3 | 22.3M/41.5M [00:00<00:00, 36.6MB/s]
- 62%|######2 | 25.9M/41.5M [00:00<00:00, 32.8MB/s]
- 77%|#######7 | 32.0M/41.5M [00:00<00:00, 37.5MB/s]
- 90%|########9 | 37.2M/41.5M [00:01<00:00, 41.2MB/s]
-100%|#########9| 41.3M/41.5M [00:01<00:00, 40.1MB/s]
-100%|##########| 41.5M/41.5M [00:01<00:00, 37.7MB/s]
+ 15%|#5 | 6.33M/41.5M [00:00<00:00, 60.2MB/s]
+ 29%|##9 | 12.1M/41.5M [00:00<00:00, 53.0MB/s]
+ 41%|####1 | 17.2M/41.5M [00:00<00:00, 32.9MB/s]
+ 58%|#####7 | 24.0M/41.5M [00:00<00:00, 39.3MB/s]
+ 77%|#######7 | 32.0M/41.5M [00:00<00:00, 47.2MB/s]
+ 92%|#########2| 38.3M/41.5M [00:00<00:00, 48.6MB/s]
+100%|##########| 41.5M/41.5M [00:00<00:00, 43.9MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_paddle.html b/docs/how_to/compile_models/from_paddle.html
index cfb016a3c2..cd86dc8824 100644
--- a/docs/how_to/compile_models/from_paddle.html
+++ b/docs/how_to/compile_models/from_paddle.html
@@ -489,7 +489,6 @@ To begin, we’ll install PaddlePaddle>=2.1.3:</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>TVM prediction top-1 id: 282, class name: 282: 'tiger cat',
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 1.022 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-paddle-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/16269b77359771348d507395692524cf/from_paddle.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">from_paddle.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index bd044335a6..fd597b1e3a 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -437,15 +437,15 @@ 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]
- 14%|#4 | 6.30M/44.7M [00:00<00:00, 48.1MB/s]
- 24%|##4 | 10.9M/44.7M [00:00<00:00, 39.6MB/s]
- 36%|###5 | 16.0M/44.7M [00:00<00:00, 32.4MB/s]
- 54%|#####3 | 24.0M/44.7M [00:00<00:00, 41.9MB/s]
- 63%|######3 | 28.2M/44.7M [00:00<00:00, 42.1MB/s]
- 72%|#######2 | 32.3M/44.7M [00:01<00:00, 22.0MB/s]
- 86%|########5 | 38.3M/44.7M [00:01<00:00, 26.1MB/s]
- 97%|#########6| 43.1M/44.7M [00:01<00:00, 30.5MB/s]
-100%|##########| 44.7M/44.7M [00:01<00:00, 30.4MB/s]
+ 14%|#4 | 6.30M/44.7M [00:00<00:00, 42.1MB/s]
+ 23%|##3 | 10.3M/44.7M [00:00<00:01, 32.5MB/s]
+ 32%|###2 | 14.3M/44.7M [00:00<00:01, 29.8MB/s]
+ 38%|###8 | 17.2M/44.7M [00:00<00:00, 29.3MB/s]
+ 54%|#####3 | 24.0M/44.7M [00:00<00:00, 38.6MB/s]
+ 68%|######7 | 30.3M/44.7M [00:00<00:00, 35.6MB/s]
+ 76%|#######5 | 33.8M/44.7M [00:01<00:00, 33.8MB/s]
+ 90%|########9 | 40.0M/44.7M [00:01<00:00, 36.9MB/s]
+100%|##########| 44.7M/44.7M [00:01<00:00, 38.4MB/s]
</pre></div>
</div>
</div>
diff --git a/docs/how_to/compile_models/from_tensorflow.html b/docs/how_to/compile_models/from_tensorflow.html
index 2d2485120b..27864b6ba8 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -657,7 +657,7 @@ 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 36.123 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 33.495 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 9e391f2290..fb09d6ef5b 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>07:18.513</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>07:06.376</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 81%" />
@@ -354,43 +354,43 @@
</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>01:36.123</p></td>
+<td><p>01:33.495</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>01:27.616</p></td>
+<td><p>01:24.849</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>01:01.022</p></td>
+<td><p>00:58.912</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>00:40.162</p></td>
+<td><p>00:39.303</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>00:34.765</p></td>
+<td><p>00:33.564</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>00:31.741</p></td>
+<td><p>00:31.068</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>00:30.066</p></td>
+<td><p>00:29.804</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>00:28.300</p></td>
+<td><p>00:27.221</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>00:25.827</p></td>
+<td><p>00:25.335</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>00:02.891</p></td>
+<td><p>00:02.824</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 3285e39c53..5552168b50 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno.html
@@ -835,10 +835,9 @@ Top5 predictions:
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 2855.3580 2855.7379 2858.1845 2852.2353 1.7204
+ 2539.4086 2539.2871 2541.9247 2538.0618 1.0440
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 0.147 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_adreno_tvmc.html b/docs/how_to/deploy_models/deploy_model_on_adreno_tvmc.html
index fcb8171744..9314935333 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno_tvmc.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno_tvmc.html
@@ -443,30 +443,33 @@ to run this tutorial with a real device over rpc.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels.h5
8192/102967424 [..............................] - ETA: 0s
- 5906432/102967424 [>.............................] - ETA: 0s
- 8380416/102967424 [=>............................] - ETA: 1s
+ 131072/102967424 [..............................] - ETA: 1:12
+ 6635520/102967424 [>.............................] - ETA: 2s
+ 8380416/102967424 [=>............................] - ETA: 3s
15024128/102967424 [===>..........................] - ETA: 1s
- 16769024/102967424 [===>..........................] - ETA: 1s
- 23412736/102967424 [=====>........................] - ETA: 1s
+ 16769024/102967424 [===>..........................] - ETA: 2s
+ 16908288/102967424 [===>..........................] - ETA: 2s
+ 23412736/102967424 [=====>........................] - ETA: 2s
25157632/102967424 [======>.......................] - ETA: 2s
+ 31203328/102967424 [========>.....................] - ETA: 1s
33546240/102967424 [========>.....................] - ETA: 1s
- 40189952/102967424 [==========>...................] - ETA: 2s
- 41934848/102967424 [===========>..................] - ETA: 2s
+ 40026112/102967424 [==========>...................] - ETA: 1s
+ 41934848/102967424 [===========>..................] - ETA: 1s
+ 48578560/102967424 [=============>................] - ETA: 1s
50323456/102967424 [=============>................] - ETA: 1s
- 53395456/102967424 [==============>...............] - ETA: 1s
56967168/102967424 [===============>..............] - ETA: 1s
58712064/102967424 [================>.............] - ETA: 1s
- 65355776/102967424 [==================>...........] - ETA: 1s
- 67100672/102967424 [==================>...........] - ETA: 1s
- 69296128/102967424 [===================>..........] - ETA: 1s
- 73744384/102967424 [====================>.........] - ETA: 0s
+ 65912832/102967424 [==================>...........] - ETA: 0s
+ 67100672/102967424 [==================>...........] - ETA: 0s
+ 69296128/102967424 [===================>..........] - ETA: 0s
+ 74522624/102967424 [====================>.........] - ETA: 0s
75489280/102967424 [====================>.........] - ETA: 0s
82124800/102967424 [======================>.......] - ETA: 0s
83877888/102967424 [=======================>......] - ETA: 0s
90521600/102967424 [=========================>....] - ETA: 0s
92266496/102967424 [=========================>....] - ETA: 0s
98910208/102967424 [===========================>..] - ETA: 0s
-100646912/102967424 [============================>.] - ETA: 0s
+100794368/102967424 [============================>.] - ETA: 0s
102850560/102967424 [============================>.] - ETA: 0s
102967424/102967424 [==============================] - 3s 0us/step
</pre></div>
diff --git a/docs/how_to/deploy_models/deploy_model_on_android.html b/docs/how_to/deploy_models/deploy_model_on_android.html
index 734ad2da68..e24272408c 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -667,7 +667,7 @@ to the remote android device.</p>
Evaluate inference time cost...
Execution time summary:
mean (ms) median (ms) max (ms) min (ms) std (ms)
- 16.8965 17.1268 17.2892 16.0530 0.4442
+ 15.8141 15.8113 15.9127 15.7327 0.0569
</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 66044e1704..6a5802bc7e 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -459,37 +459,38 @@ 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]
- 4%|3 | 6.30M/170M [00:00<00:03, 56.2MB/s]
- 7%|6 | 11.7M/170M [00:00<00:03, 46.6MB/s]
- 10%|9 | 16.2M/170M [00:00<00:05, 29.1MB/s]
- 13%|#3 | 22.3M/170M [00:00<00:04, 37.4MB/s]
- 16%|#5 | 26.5M/170M [00:00<00:04, 32.2MB/s]
- 19%|#8 | 32.0M/170M [00:01<00:05, 27.6MB/s]
- 24%|##3 | 40.0M/170M [00:01<00:03, 36.4MB/s]
- 27%|##7 | 46.3M/170M [00:01<00:03, 39.3MB/s]
- 30%|##9 | 50.5M/170M [00:01<00:03, 32.7MB/s]
- 33%|###2 | 56.0M/170M [00:01<00:03, 34.8MB/s]
- 38%|###7 | 64.0M/170M [00:01<00:02, 38.8MB/s]
- 42%|####2 | 72.0M/170M [00:02<00:02, 43.8MB/s]
- 46%|####6 | 78.3M/170M [00:02<00:02, 40.3MB/s]
- 48%|####8 | 82.3M/170M [00:02<00:02, 37.0MB/s]
- 51%|##### | 86.3M/170M [00:02<00:02, 33.4MB/s]
- 53%|#####2 | 89.6M/170M [00:02<00:02, 32.8MB/s]
- 56%|#####5 | 94.3M/170M [00:02<00:02, 34.1MB/s]
- 57%|#####7 | 97.6M/170M [00:02<00:02, 33.2MB/s]
- 60%|###### | 102M/170M [00:03<00:02, 32.4MB/s]
- 62%|######2 | 105M/170M [00:03<00:02, 30.1MB/s]
- 66%|######5 | 112M/170M [00:03<00:01, 34.3MB/s]
- 71%|####### | 120M/170M [00:03<00:01, 43.3MB/s]
- 75%|#######5 | 128M/170M [00:03<00:01, 41.8MB/s]
- 80%|######## | 136M/170M [00:03<00:00, 38.2MB/s]
- 85%|########4 | 144M/170M [00:04<00:00, 40.0MB/s]
- 88%|########8 | 150M/170M [00:04<00:00, 43.5MB/s]
- 91%|#########1| 155M/170M [00:04<00:00, 35.2MB/s]
- 93%|#########3| 158M/170M [00:04<00:00, 30.9MB/s]
- 95%|#########5| 162M/170M [00:04<00:00, 28.5MB/s]
- 98%|#########7| 166M/170M [00:04<00:00, 30.6MB/s]
-100%|##########| 170M/170M [00:04<00:00, 35.6MB/s]
+ 4%|3 | 6.30M/170M [00:00<00:07, 23.3MB/s]
+ 5%|5 | 8.53M/170M [00:00<00:07, 22.9MB/s]
+ 8%|8 | 14.3M/170M [00:00<00:09, 17.4MB/s]
+ 11%|# | 18.5M/170M [00:00<00:07, 22.3MB/s]
+ 14%|#4 | 24.0M/170M [00:01<00:05, 26.0MB/s]
+ 19%|#8 | 32.0M/170M [00:01<00:04, 31.4MB/s]
+ 23%|##2 | 38.6M/170M [00:01<00:03, 38.6MB/s]
+ 25%|##5 | 42.9M/170M [00:01<00:04, 31.9MB/s]
+ 28%|##8 | 48.0M/170M [00:01<00:03, 32.1MB/s]
+ 33%|###2 | 56.0M/170M [00:01<00:03, 35.4MB/s]
+ 38%|###7 | 64.0M/170M [00:02<00:02, 40.3MB/s]
+ 42%|####2 | 72.0M/170M [00:02<00:02, 45.3MB/s]
+ 46%|####6 | 78.3M/170M [00:02<00:02, 41.9MB/s]
+ 49%|####8 | 82.5M/170M [00:02<00:02, 36.8MB/s]
+ 52%|#####1 | 88.0M/170M [00:02<00:02, 34.5MB/s]
+ 56%|#####5 | 94.3M/170M [00:02<00:02, 36.6MB/s]
+ 58%|#####7 | 97.9M/170M [00:03<00:02, 35.6MB/s]
+ 60%|###### | 102M/170M [00:03<00:02, 35.0MB/s]
+ 62%|######2 | 106M/170M [00:03<00:02, 33.6MB/s]
+ 66%|######5 | 112M/170M [00:03<00:01, 35.1MB/s]
+ 71%|####### | 120M/170M [00:03<00:01, 38.2MB/s]
+ 74%|#######4 | 126M/170M [00:03<00:01, 42.7MB/s]
+ 77%|#######6 | 130M/170M [00:03<00:00, 42.3MB/s]
+ 80%|######## | 136M/170M [00:04<00:00, 38.8MB/s]
+ 85%|########4 | 144M/170M [00:04<00:00, 42.1MB/s]
+ 88%|########8 | 150M/170M [00:04<00:00, 42.9MB/s]
+ 91%|######### | 154M/170M [00:04<00:00, 40.1MB/s]
+ 93%|#########3| 158M/170M [00:04<00:00, 39.7MB/s]
+ 95%|#########5| 162M/170M [00:04<00:00, 33.7MB/s]
+ 98%|#########7| 166M/170M [00:05<00:00, 31.1MB/s]
+100%|#########9| 169M/170M [00:05<00:00, 30.9MB/s]
+100%|##########| 170M/170M [00:05<00:00, 34.8MB/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=& [...]
@@ -587,7 +588,7 @@ 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 52.109 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes 41.966 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 f7cfcec6bd..a296e6e6b4 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -500,9 +500,8 @@ 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, 37.1MB/s]
- 90%|########9 | 12.2M/13.6M [00:00<00:00, 38.0MB/s]
-100%|##########| 13.6M/13.6M [00:00<00:00, 41.4MB/s]
+ 59%|#####8 | 7.99M/13.6M [00:00<00:00, 73.5MB/s]
+100%|##########| 13.6M/13.6M [00:00<00:00, 107MB/s]
</pre></div>
</div>
</div>
@@ -593,7 +592,7 @@ 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)
- 90.6979 90.5879 95.3150 90.3424 0.5525
+ 90.6953 90.6578 91.6423 90.3184 0.2578
</pre></div>
</div>
<div class="admonition note">
@@ -632,7 +631,7 @@ 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 21.891 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 18.908 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 b9281e1a65..343efa2b2a 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -585,7 +585,7 @@ 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)
- 120.5770 120.5138 122.3097 119.4893 0.4387
+ 120.9231 120.8937 121.9398 120.0892 0.3639
</pre></div>
</div>
<div class="admonition note">
@@ -613,7 +613,7 @@ 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 32.983 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes 40.736 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 52bd310c1c..4b2bf1aec6 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -526,7 +526,7 @@ 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 56.443 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 52.802 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 4d713f4ca8..8703a27152 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -468,26 +468,24 @@ 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 | 3891/132723 [00:00<00:03, 38706.09KB/s]
- 9%|8 | 11899/132723 [00:00<00:01, 62987.25KB/s]
- 14%|#3 | 18203/132723 [00:00<00:02, 46444.99KB/s]
- 20%|## | 26621/132723 [00:00<00:01, 58685.28KB/s]
- 25%|##4 | 32995/132723 [00:00<00:02, 43196.03KB/s]
- 31%|###1 | 41420/132723 [00:00<00:01, 53244.53KB/s]
- 37%|###7 | 49140/132723 [00:01<00:01, 45672.94KB/s]
- 43%|####3 | 57536/132723 [00:01<00:01, 54148.48KB/s]
- 49%|####9 | 65522/132723 [00:01<00:01, 46448.10KB/s]
- 56%|#####5 | 73882/132723 [00:01<00:01, 54213.92KB/s]
- 61%|###### | 80692/132723 [00:01<00:01, 47423.74KB/s]
- 65%|######4 | 86191/132723 [00:01<00:01, 42013.67KB/s]
- 69%|######8 | 90955/132723 [00:01<00:00, 42956.29KB/s]
- 74%|#######3 | 97594/132723 [00:02<00:00, 42894.85KB/s]
- 77%|#######6 | 102186/132723 [00:02<00:00, 34811.40KB/s]
- 83%|########3 | 110653/132723 [00:02<00:00, 44960.92KB/s]
- 87%|########7 | 115900/132723 [00:02<00:00, 35105.08KB/s]
- 94%|#########3| 124342/132723 [00:02<00:00, 44718.36KB/s]
-100%|##########| 132723/132723 [00:02<00:00, 53157.78KB/s]
-100%|##########| 132723/132723 [00:02<00:00, 46862.48KB/s]
+ 5%|5 | 7222/132723 [00:00<00:01, 72208.53KB/s]
+ 12%|#1 | 15768/132723 [00:00<00:01, 79994.41KB/s]
+ 18%|#7 | 23768/132723 [00:00<00:01, 76680.01KB/s]
+ 24%|##4 | 32330/132723 [00:00<00:01, 80117.08KB/s]
+ 30%|### | 40361/132723 [00:00<00:01, 74549.15KB/s]
+ 37%|###6 | 48855/132723 [00:00<00:01, 77872.28KB/s]
+ 43%|####2 | 56707/132723 [00:00<00:01, 61594.84KB/s]
+ 48%|####8 | 64310/132723 [00:00<00:01, 60647.30KB/s]
+ 53%|#####3 | 70722/132723 [00:01<00:01, 51342.05KB/s]
+ 60%|#####9 | 79226/132723 [00:01<00:00, 59245.91KB/s]
+ 65%|######4 | 85680/132723 [00:01<00:00, 54381.58KB/s]
+ 71%|####### | 94201/132723 [00:01<00:00, 61887.88KB/s]
+ 76%|#######5 | 100855/132723 [00:01<00:00, 45686.07KB/s]
+ 80%|######## | 106307/132723 [00:01<00:00, 43095.95KB/s]
+ 86%|########6 | 114680/132723 [00:02<00:00, 43513.34KB/s]
+ 92%|#########1| 121550/132723 [00:02<00:00, 48665.68KB/s]
+ 98%|#########8| 130086/132723 [00:02<00:00, 56963.55KB/s]
+100%|##########| 132723/132723 [00:02<00:00, 58034.67KB/s]
</pre></div>
</div>
<p>Create TVM runtime and do inference
@@ -526,7 +524,7 @@ 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> ( 4 minutes 3.870 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> ( 4 minutes 4.065 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 a6aa3ea1d9..2e2a906351 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>17:28.952</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>17:11.643</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 86%" />
@@ -354,43 +354,43 @@
</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>04:03.870</p></td>
+<td><p>04:04.065</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>03:52.109</p></td>
+<td><p>03:41.966</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>02:32.983</p></td>
+<td><p>02:40.736</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>01:56.443</p></td>
+<td><p>01:52.802</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>01:21.891</p></td>
+<td><p>01:18.908</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>01:00.147</p></td>
+<td><p>00:55.492</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_adreno_tvmc.html#sphx-glr-how-to-deploy-models-deploy-model-on-adreno-tvmc-py"><span class="std std-ref">Deploy the Pretrained Model on Adreno™ with tvmc Interface</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_adreno_tvmc.py</span></code>)</p></td>
-<td><p>00:56.226</p></td>
+<td><p>00:54.722</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>00:46.295</p></td>
+<td><p>00:44.215</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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>00:29.795</p></td>
+<td><p>00:29.579</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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>00:29.187</p></td>
+<td><p>00:29.152</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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 fec6a4d507..c2882b4dc1 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -624,7 +624,7 @@ 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.zip0c889e00-21d9-47ea-ad4b-37a33656e572 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.zip7a614ddb-3585-4d6c-bc26-b24d2b3b1ce4 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 14c82c6fcd..ff4b3a65f3 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>00:57.089</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:55.739</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -354,15 +354,15 @@
</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>00:52.825</p></td>
+<td><p>00:51.813</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>00:03.131</p></td>
+<td><p>00:02.822</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>00:01.125</p></td>
+<td><p>00:01.097</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 e280b1c0d5..c1dbc27e72 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -531,10 +531,10 @@ 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: 23251us [23251us] (48.99%; 48.99%)
-FoldScaleAxis: 24207us [9us] (51.01%; 51.01%)
- FoldConstant: 24198us [1733us] (50.99%; 99.96%)
- InferType: 22466us [22466us] (47.34%; 92.84%)
+InferType: 22959us [22959us] (48.77%; 48.77%)
+FoldScaleAxis: 24120us [7us] (51.23%; 51.23%)
+ FoldConstant: 24112us [1699us] (51.22%; 99.97%)
+ InferType: 22414us [22414us] (47.61%; 92.95%)
</pre></div>
</div>
</div>
@@ -556,10 +556,10 @@ 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: 22506us [22506us] (48.06%; 48.06%)
-FoldScaleAxis: 24319us [7us] (51.94%; 51.94%)
- FoldConstant: 24312us [1759us] (51.92%; 99.97%)
- InferType: 22553us [22553us] (48.16%; 92.76%)
+InferType: 22696us [22696us] (48.29%; 48.29%)
+FoldScaleAxis: 24300us [7us] (51.71%; 51.71%)
+ FoldConstant: 24294us [1739us] (51.69%; 99.97%)
+ InferType: 22554us [22554us] (47.99%; 92.84%)
</pre></div>
</div>
<p>Register empty list to clear existing instruments.</p>
diff --git a/docs/how_to/optimize_operators/opt_conv_cuda.html b/docs/how_to/optimize_operators/opt_conv_cuda.html
index 1f8a090ddd..aeb334510f 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -580,7 +580,7 @@ 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: 53.514942 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 38.739711 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 321d35b256..5dea2f61bf 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -862,7 +862,7 @@ 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: 12.233948 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 12.234931 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 24d68660b7..f93abfeaa2 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -477,8 +477,8 @@ 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: 0.018487
-Baseline: 3.292571
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.017871
+Baseline: 3.441978
</pre></div>
</div>
<p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -537,7 +537,7 @@ 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: 0.310587
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.300670
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -594,7 +594,7 @@ 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: 0.343112
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.329739
</pre></div>
</div>
<p>Here is the generated IR after vectorization.</p>
@@ -649,7 +649,7 @@ 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: 0.115167
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.114937
</pre></div>
</div>
<p>Here is the generated IR after loop permutation.</p>
@@ -726,7 +726,7 @@ 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: 0.109182
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.110421
</pre></div>
</div>
<p>Here is the generated IR after array packing.</p>
@@ -804,7 +804,7 @@ 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: 0.110476
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.114298
</pre></div>
</div>
<p>Here is the generated IR after blocking.</p>
@@ -884,7 +884,7 @@ 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: 0.146319
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.147723
</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 82bbeca1db..b3b8369a38 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>00:35.166</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:35.429</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -354,15 +354,15 @@
</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>00:32.177</p></td>
+<td><p>00:32.450</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>00:01.862</p></td>
+<td><p>00:01.904</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>00:01.127</p></td>
+<td><p>00:01.076</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 d73ca78b7c..88007abfe3 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>10:38.424</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>10:41.295</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 85%" />
@@ -354,27 +354,27 @@
</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>06:38.700</p></td>
+<td><p>06:27.997</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>01:44.702</p></td>
+<td><p>01:46.055</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>01:12.714</p></td>
+<td><p>01:12.916</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>00:33.748</p></td>
+<td><p>00:45.647</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>00:14.574</p></td>
+<td><p>00:14.723</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>00:13.986</p></td>
+<td><p>00:13.958</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 319b6b96ce..46a98a907c 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
@@ -512,791 +512,518 @@ class Module:
T.func_attr({"from_legacy_te_schedule": T.bool(True), "global_symbol": "main", "tir.noalias": T.bool(True)})
blockIdx_x = T.launch_thread("blockIdx.x", 32)
conv2d_nchw = T.allocate([14], "float32", "local")
- pad_temp_shared = T.allocate([6272], "float32", "shared")
- kernel_shared = T.allocate([2048], "float32", "shared")
+ pad_temp_shared = T.allocate([4032], "float32", "shared")
+ kernel_shared = T.allocate([3072], "float32", "shared")
threadIdx_x = T.launch_thread("threadIdx.x", 56)
- conv2d_nchw_1 = T.Buffer((4,), data=conv2d_nchw, scope="local", align=8)
+ conv2d_nchw_1 = T.Buffer((49,), data=conv2d_nchw, scope="local", align=16)
conv2d_nchw_1[0] = T.float32(0)
- conv2d_nchw_1[2] = T.float32(0)
- conv2d_nchw_1[4] = T.float32(0)
- conv2d_nchw_1[6] = T.float32(0)
- conv2d_nchw_1[8] = T.float32(0)
- conv2d_nchw_1[10] = T.float32(0)
- conv2d_nchw_1[12] = T.float32(0)
- conv2d_nchw_1[1] = T.float32(0)
- conv2d_nchw_1[3] = T.float32(0)
- conv2d_nchw_1[5] = T.float32(0)
conv2d_nchw_1[7] = T.float32(0)
+ conv2d_nchw_1[1] = T.float32(0)
+ conv2d_nchw_1[8] = T.float32(0)
+ conv2d_nchw_1[2] = T.float32(0)
conv2d_nchw_1[9] = T.float32(0)
+ conv2d_nchw_1[3] = T.float32(0)
+ conv2d_nchw_1[10] = T.float32(0)
+ conv2d_nchw_1[4] = T.float32(0)
conv2d_nchw_1[11] = T.float32(0)
+ conv2d_nchw_1[5] = T.float32(0)
+ conv2d_nchw_1[12] = T.float32(0)
+ conv2d_nchw_1[6] = T.float32(0)
conv2d_nchw_1[13] = T.float32(0)
- for rc_outer_outer, ry_outer_outer, rx_outer_outer in T.grid(4, 3, 3):
- cse_var_2: T.int32 = rc_outer_outer * 1152
- cse_var_1: T.int32 = ry_outer_outer * 3
+ for rc_outer_outer, rx_outer_outer in T.grid(8, 3):
+ cse_var_2: T.int32 = rc_outer_outer * 3136
+ cse_var_1: T.int32 = rc_outer_outer * 576
threadIdx_x_1 = T.env_thread("threadIdx.x")
- pad_temp_shared_1 = T.Buffer((6272,), data=pad_temp_shared, scope="shared")
+ pad_temp_shared_1 = T.Buffer((4032,), data=pad_temp_shared, scope="shared")
data_1 = T.Buffer((25088,), data=data.data)
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer - 8], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 48], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 104], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 168] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 160], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 216], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 280] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 272], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 328], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 392] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 384], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 440], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 504] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 496], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 552], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 616] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 608], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 664], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 728] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 720], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 776], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 840] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 832], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 888], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 952] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 944], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1000], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1064] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1056], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1112], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1176] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1168], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1224], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1288] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1280], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1344] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1336], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1400] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1392], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1456] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1448], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1512] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1504], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1568] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1560], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1624] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1616], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1680] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1672], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1736] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1728], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 56] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 56) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1792] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1784], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 112] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 112) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1848] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1840], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 168] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 168) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1904] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1896], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 224] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 224) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 1960] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 1952], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 280] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 280) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2016] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2008], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 336] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 336) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2072] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2064], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 392] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 392) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2128] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2120], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 448] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 448) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2184] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2176], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 504] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 384], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2240] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2232], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 560] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 560) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2296] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2288], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 616] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 616) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2352] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2344], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 672] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 672) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2408] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2400], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 728] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 728) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2464] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2456], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 784] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 784) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2520] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2512], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 840] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 840) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2576] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2568], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 896] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 896) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2632] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2624], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 952] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 952) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2688] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2680], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1008] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 776], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2744] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2736], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1064] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1064) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2800] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2792], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1120] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1120) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2856] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2848], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1176] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1176) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2912] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2904], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1232] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1232) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 2968] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 2960], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1288] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1288) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3024] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3016], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1344] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1344) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3080] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3072], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1400] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1400) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3136] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3128], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1456] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1456) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3192] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3184], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1512] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1168], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3248] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3240], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1568] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1568) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3304] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3296], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1624] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1624) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3360] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3352], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1680] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1680) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3416] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3408], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1736] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1736) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3472] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3464], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1792] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1792) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3528] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3520], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1848] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1848) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3584] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3576], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1904] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1904) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3640] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3632], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 1960] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 1960) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3696] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3688], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2016] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1560], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3752] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3744], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2072] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2072) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3808] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3800], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2128] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2128) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3864] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3856], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2184] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2184) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3920] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3912], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2240] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2240) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 3976] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 3968], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2296] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2296) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4032] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4024], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2352] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2352) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4088] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4080], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2408] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2408) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4144] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4136], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2464] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2464) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4200] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4192], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2520] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 1952], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4256] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4248], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2576] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2576) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4312] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4304], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2632] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2632) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4368] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4360], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2688] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2688) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4424] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4416], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2744] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2744) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4480] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4472], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2800] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2800) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4536] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4528], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2856] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2856) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4592] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4584], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2912] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2912) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4648] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4640], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 2968] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 2968) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4704] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4696], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3024] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 2344], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4760] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4752], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3080] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3080) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4816] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4808], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3136] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3136) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4872] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4864], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3192] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3192) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4928] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4920], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3248] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3248) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 4984] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 4976], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3304] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3304) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5040] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5032], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3360] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3360) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5096] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5088], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3416] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3416) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5152] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5144], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3472] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3472) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5208] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5200], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3528] = T.if_then_else(7 <= threadIdx_x_1 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + threadIdx_x_1 + rx_outer_outer + 2736], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5264] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5256], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3584] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 8) % 9 and (threadIdx_x_1 // 7 + 8) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3584) // 63 * 49 + (threadIdx_x_1 // 7 + 8) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5320] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5312], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3640] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 7) % 9 and (threadIdx_x_1 // 7 + 7) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3640) // 63 * 49 + (threadIdx_x_1 // 7 + 7) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5376] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5368], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3696] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 6) % 9 and (threadIdx_x_1 // 7 + 6) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3696) // 63 * 49 + (threadIdx_x_1 // 7 + 6) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5432] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5424], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3752] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 5) % 9 and (threadIdx_x_1 // 7 + 5) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3752) // 63 * 49 + (threadIdx_x_1 // 7 + 5) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5488] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5480], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3808] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 4) % 9 and (threadIdx_x_1 // 7 + 4) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3808) // 63 * 49 + (threadIdx_x_1 // 7 + 4) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5544] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5536], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3864] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 3) % 9 and (threadIdx_x_1 // 7 + 3) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3864) // 63 * 49 + (threadIdx_x_1 // 7 + 3) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5600] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5592], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3920] = T.if_then_else(1 <= (threadIdx_x_1 // 7 + 2) % 9 and (threadIdx_x_1 // 7 + 2) % 9 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3920) // 63 * 49 + (threadIdx_x_1 // 7 + 2) % 9 * 7 + rx_outer_outer + threadIdx_x_1 % 7 - 8], T.float32(0))
with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5656] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5648], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5712] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5704], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5768] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5760], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5824] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5816], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5880] = T.if_then_else(1 <= threadIdx_x_1 % 49 // 7 + ry_outer_outer and threadIdx_x_1 % 49 // 7 + ry_outer_outer < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5872], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5936] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 1) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5928], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 5992] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 2) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 5984], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6048] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 3) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6040], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6104] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 4) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6096], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6160] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 5) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6152], T.float32(0))
- with T.launch_thread(threadIdx_x_1, 56):
- pad_temp_shared_1[threadIdx_x_1 + 6216] = T.if_then_else(1 <= ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 and ry_outer_outer + (threadIdx_x_1 // 7 + 6) % 7 < 8 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[rc_outer_outer * 6272 + ry_outer_outer * 7 + threadIdx_x_1 + rx_outer_outer + 6208], T.float32(0))
+ pad_temp_shared_1[threadIdx_x_1 + 3976] = T.if_then_else(threadIdx_x_1 < 49 and 1 <= rx_outer_outer + threadIdx_x_1 % 7 and rx_outer_outer + threadIdx_x_1 % 7 < 8, data_1[cse_var_2 + (threadIdx_x_1 + 3976) // 63 * 49 + threadIdx_x_1 + rx_outer_outer - 1], T.float32(0))
threadIdx_x_2 = T.env_thread("threadIdx.x")
- kernel_shared_1 = T.Buffer((2048,), data=kernel_shared, scope="shared")
+ kernel_shared_1 = T.Buffer((3072,), data=kernel_shared, scope="shared")
kernel_1 = T.Buffer((2359296,), data=kernel.data)
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 168) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 224) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 280) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 336) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 144) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 392) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 24]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 448) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 192]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 504) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 360]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 560) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 176) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 616) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 120]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 672) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 288]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 728) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 152) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 784) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 48]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 56] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 840) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 216]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 112] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 112) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 896) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 384]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 168] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 168) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 952) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 184) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 224] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 224) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 96) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1008) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 144]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 280] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 280) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 216]
+ kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1064) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 312]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 336] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 336) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 80) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1120) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 160) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 392] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 392) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 72]
+ kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1176) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 72]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 448] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 448) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 576]
+ kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1232) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 240]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 504] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 504) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 120) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1288) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 408]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 560] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 560) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 432]
+ kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 32256]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 616] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 616) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 104) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1400) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 672] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 672) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 288]
+ kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1456) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 728] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 728) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 88) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1512) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 784] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 784) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 144]
+ kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1568) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 840] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 840) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 648]
+ kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1624) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 896] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 32256]
+ kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1680) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 144) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 952] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 952) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1736) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 24]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1008] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1008) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1792) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 192]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1064] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1064) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1848) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 360]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1120] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1120) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 96) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1904) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 176) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1176] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1176) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 216]
+ kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1960) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 120]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1232] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1232) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 80) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2016) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 288]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1288] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1288) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 72]
+ kernel_shared_1[threadIdx_x_2 + 2072] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2072) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 152) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1344] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1344) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 576]
+ kernel_shared_1[threadIdx_x_2 + 2128] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2128) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 48]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1400] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1400) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 120) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2184] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2184) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 216]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1456] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1456) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 432]
+ kernel_shared_1[threadIdx_x_2 + 2240] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2240) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 384]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1512] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1512) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 104) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2296] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2296) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 184) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1568] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1568) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 288]
+ kernel_shared_1[threadIdx_x_2 + 2352] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2352) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 144]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1624] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1624) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 88) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2408] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2408) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 312]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1680] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1680) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 144]
+ kernel_shared_1[threadIdx_x_2 + 2464] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2464) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 160) % 192 * 3 + rx_outer_outer]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1736] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1736) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 648]
+ kernel_shared_1[threadIdx_x_2 + 2520] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2520) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 72]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1792] = kernel_1[blockIdx_x * 73728 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 64512]
+ kernel_shared_1[threadIdx_x_2 + 2576] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2576) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 240]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1848] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1848) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 504]
+ kernel_shared_1[threadIdx_x_2 + 2632] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2632) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 408]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1904] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1904) // 128 * 4608 + cse_var_2 + (threadIdx_x_2 + 112) % 128 * 9 + cse_var_1 + rx_outer_outer]
+ kernel_shared_1[threadIdx_x_2 + 2688] = kernel_1[blockIdx_x * 73728 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 64512]
with T.launch_thread(threadIdx_x_2, 56):
- kernel_shared_1[threadIdx_x_2 + 1960] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 1960) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 360]
+ kernel_shared_1[threadIdx_x_2 + 2744] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2744) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 168]
with T.launch_thread(threadIdx_x_2, 56):
- if T.likely(threadIdx_x_2 < 32):
- kernel_shared_1[threadIdx_x_2 + 2016] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2016) // 128 * 4608 + cse_var_2 + threadIdx_x_2 * 9 + cse_var_1 + rx_outer_outer + 864]
- for rc_outer_inner in range(4):
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 128]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 49] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 50] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 51] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 52] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 53] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 1]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 49] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 50] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 51] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 52] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 53] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 54] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 55] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 129]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 98] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 102] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 103] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 104] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 2]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 98] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 99] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 100] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 101] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 102] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 103] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 104] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 130]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 147] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 148] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 149] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 150] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 151] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 152] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 153] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 3]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 147] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 148] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 149] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 150] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 151] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 152] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 153] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 131]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 196] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 197] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 198] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 199] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 200] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 201] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 202] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 4]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 196] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 197] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 198] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 199] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 200] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 201] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 202] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 132]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 245] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 246] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 247] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 248] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 249] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 250] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 251] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 5]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 245] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 246] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 247] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 248] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 249] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 250] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 251] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 133]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 294] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 295] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 296] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 297] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 298] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 299] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 300] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 6]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 294] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 295] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 296] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 297] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 298] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 299] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 300] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 134]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 343] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 344] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 345] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 346] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 347] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 348] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 349] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 7]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 343] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 344] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 345] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 346] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 347] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 348] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 349] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 135]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 392] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 393] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 394] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 395] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 396] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 397] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 398] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 8]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 392] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 393] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 394] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 395] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 396] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 397] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 398] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 136]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 9]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 137]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 490] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 491] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 492] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 493] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 494] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 495] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 496] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 10]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 490] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 491] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 492] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 493] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 494] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 495] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 496] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 138]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 539] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 540] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 541] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 542] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 543] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 544] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 545] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 11]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 539] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 540] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 541] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 542] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 543] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 544] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 545] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 139]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 588] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 589] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 590] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 591] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 592] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 593] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 594] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 12]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 588] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 589] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 590] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 591] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 592] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 593] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 594] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 140]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 637] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 638] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 639] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 640] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 641] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 642] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 643] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 13]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 637] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 638] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 639] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 640] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 641] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 642] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 643] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 141]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 686] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 687] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 688] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 689] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 690] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 691] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 692] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 14]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 686] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 687] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 688] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 689] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 690] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 691] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 692] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 142]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 735] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 736] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 737] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 738] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 739] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 740] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 741] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 15]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 735] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 736] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 737] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 738] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 739] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 740] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 741] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 143]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 784] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 785] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 786] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 787] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 788] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 789] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 790] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 16]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 784] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 785] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 786] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 787] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 788] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 789] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 790] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 144]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 833] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 834] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 835] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 836] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 837] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 838] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 839] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 17]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 833] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 834] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 835] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 836] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 837] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 838] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 839] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 145]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 18]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 146]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 931] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 932] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 933] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 934] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 935] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 936] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 937] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 19]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 931] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 932] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 933] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 934] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 935] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 936] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 937] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 147]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 980] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 981] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 982] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 983] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 984] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 985] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 986] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 20]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 980] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 981] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 982] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 983] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 984] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 985] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 986] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 148]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1029] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1030] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1031] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1032] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1033] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1034] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1035] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 21]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1029] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1030] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1031] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1032] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1033] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1034] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1035] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 149]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1078] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1079] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1080] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1081] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1082] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1083] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1084] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 22]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1078] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1079] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1080] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1081] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1082] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1083] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1084] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 150]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1127] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1128] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1129] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1130] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1131] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1132] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1133] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 23]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1127] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1128] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1129] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1130] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1131] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1132] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1133] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 151]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1176] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1177] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1178] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1179] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1180] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1181] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1182] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 24]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1176] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1177] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1178] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1179] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1180] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1181] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1182] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 152]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1225] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1226] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1227] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1228] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1229] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1230] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1231] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 25]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1225] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1226] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1227] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1228] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1229] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1230] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1231] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 153]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1274] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1275] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1276] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1277] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1278] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1279] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1280] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 26]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1274] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1275] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1276] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1277] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1278] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1279] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1280] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 154]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1323] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1324] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1325] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1326] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1327] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1328] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1329] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 27]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1323] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1324] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1325] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1326] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1327] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1328] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1329] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 155]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1372] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1373] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1374] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1375] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1376] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1377] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1378] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 28]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1372] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1373] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1374] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1375] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1376] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1377] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1378] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 156]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1421] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1422] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1423] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1424] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1425] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1426] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1427] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 29]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1421] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1422] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1423] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1424] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1425] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1426] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1427] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 157]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1470] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1471] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1472] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1473] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1474] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1475] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1476] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 30]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1470] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1471] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1472] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1473] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1474] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1475] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1476] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 158]
- conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1519] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1520] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1521] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1522] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1523] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1524] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1525] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 31]
- conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1519] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1520] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1521] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1522] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1523] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1524] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1568 + threadIdx_x % 7 * 7 + 1525] * kernel_shared_1[threadIdx_x // 7 * 256 + rc_outer_inner * 32 + 159]
- for i1_inner in range(2):
+ kernel_shared_1[threadIdx_x_2 + 2800] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2800) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 336]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2856] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2856) // 192 * 4608 + cse_var_1 + (threadIdx_x_2 + 168) % 192 * 3 + rx_outer_outer]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2912] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2912) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 96]
+ with T.launch_thread(threadIdx_x_2, 56):
+ kernel_shared_1[threadIdx_x_2 + 2968] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 2968) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 264]
+ with T.launch_thread(threadIdx_x_2, 56):
+ if T.likely(threadIdx_x_2 < 48):
+ kernel_shared_1[threadIdx_x_2 + 3024] = kernel_1[blockIdx_x * 73728 + (threadIdx_x_2 + 3024) // 192 * 4608 + cse_var_1 + threadIdx_x_2 * 3 + rx_outer_outer + 432]
+ for rc_outer_inner, ry_outer_inner in T.grid(4, 3):
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 63] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 126] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 126] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 189] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 189] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 252] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 252] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 315] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 315] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 378] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 378] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 441] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 504] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 504] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 567] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 567] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 630] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 630] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 693] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 693] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 756] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 756] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 819] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 819] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 882] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[0] = conv2d_nchw_1[0] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 945] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[7] = conv2d_nchw_1[7] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 945] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 1] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 64] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 127] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 127] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 190] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 190] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 253] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 253] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 316] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 316] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 379] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 379] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 442] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 505] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 505] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 568] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 568] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 631] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 631] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 694] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 694] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 757] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 757] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 820] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 820] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 883] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[1] = conv2d_nchw_1[1] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 946] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[8] = conv2d_nchw_1[8] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 946] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 2] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 65] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 128] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 128] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 191] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 191] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 254] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 254] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 317] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 317] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 380] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 380] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 443] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 506] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 506] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 569] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 569] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 632] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 632] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 695] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 695] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 758] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 758] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 821] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 821] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 884] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[2] = conv2d_nchw_1[2] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 947] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[9] = conv2d_nchw_1[9] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 947] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 3] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 66] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 66] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 129] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 129] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 192] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 192] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 255] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 255] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 318] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 318] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 381] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 381] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 444] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 507] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 507] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 570] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 570] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 633] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 633] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 696] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 696] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 759] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 759] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 822] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 822] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 885] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[3] = conv2d_nchw_1[3] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 948] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[10] = conv2d_nchw_1[10] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 948] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 4] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 67] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 67] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 130] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 130] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 193] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 193] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 256] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 256] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 319] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 319] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 382] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 382] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 445] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 508] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 508] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 571] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 571] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 634] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 634] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 697] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 697] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 760] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 760] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 823] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 823] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 886] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[4] = conv2d_nchw_1[4] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 949] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[11] = conv2d_nchw_1[11] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 949] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 5] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 68] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 68] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 131] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 131] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 194] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 194] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 257] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 257] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 320] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 320] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 383] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 383] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 446] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 509] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 509] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 572] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 572] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 635] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 635] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 698] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 698] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 761] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 761] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 824] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 824] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 887] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[5] = conv2d_nchw_1[5] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 950] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[12] = conv2d_nchw_1[12] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 950] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 6] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1536]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 69] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 3]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 69] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1539]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 132] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 6]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 132] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1542]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 195] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 9]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 195] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1545]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 258] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 12]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 258] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1548]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 321] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 15]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 321] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1551]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 384] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 18]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 384] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1554]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 21]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 447] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1557]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 510] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 24]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 510] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1560]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 573] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 27]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 573] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1563]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 636] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 30]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 636] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1566]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 699] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 33]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 699] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1569]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 762] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 36]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 762] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1572]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 825] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 39]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 825] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1575]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 42]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 888] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1578]
+ conv2d_nchw_1[6] = conv2d_nchw_1[6] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 951] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 45]
+ conv2d_nchw_1[13] = conv2d_nchw_1[13] + pad_temp_shared_1[rc_outer_inner * 1008 + ry_outer_inner * 7 + threadIdx_x % 7 * 7 + 951] * kernel_shared_1[threadIdx_x // 7 * 192 + rc_outer_inner * 48 + ry_outer_inner + 1581]
+ 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 * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7] = T.max(conv2d_nchw_1[i1_inner] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 1] = T.max(conv2d_nchw_1[i1_inner + 2] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 2] = T.max(conv2d_nchw_1[i1_inner + 4] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 3] = T.max(conv2d_nchw_1[i1_inner + 6] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 4] = T.max(conv2d_nchw_1[i1_inner + 8] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 5] = T.max(conv2d_nchw_1[i1_inner + 10] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
- compute_1[blockIdx_x * 784 + threadIdx_x // 7 * 98 + i1_inner * 49 + threadIdx_x % 7 * 7 + 6] = T.max(conv2d_nchw_1[i1_inner + 12] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 * 2 + i1_inner], T.float32(0))
+ compute_1[blockIdx_x * 784 + threadIdx_x * 7 + i3_inner] = T.max(conv2d_nchw_1[i3_inner] + bias_1[blockIdx_x * 16 + threadIdx_x // 7], T.float32(0))
+ compute_1[blockIdx_x * 784 + threadIdx_x * 7 + i3_inner + 392] = T.max(conv2d_nchw_1[i3_inner + 7] + bias_1[blockIdx_x * 16 + threadIdx_x // 7 + 8], T.float32(0))
</pre></div>
</div>
</div>
@@ -1330,7 +1057,7 @@ 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: 0.257 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.230 ms
</pre></div>
</div>
</div>
@@ -1359,37 +1086,37 @@ conv2d_nchw_nn_o_i, conv2d_nchw_nn_i = s[conv2d_nchw].split(conv2d_nchw_nn, fact
conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_i, factor=1)
conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
-conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=2)
+conv2d_nchw_ff_o_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=8)
-conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=7)
conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
-conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
+conv2d_nchw_xx_o_o_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=7)
-conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=32)
+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=16)
conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
-conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
+conv2d_nchw_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=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=2)
+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=8)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=7)
compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
-compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
+compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
-compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=7)
+compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
kernel_shared = s.cache_read(kernel, "shared", [conv2d_nchw])
@@ -1415,7 +1142,7 @@ pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fus
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=56)
s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
-s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
CUDA source code:
@@ -1442,639 +1169,388 @@ CUDA source code:
#endif
extern "C" __global__ void __launch_bounds__(56) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
float conv2d_nchw[14];
- __shared__ float pad_temp_shared[6272];
- __shared__ float kernel_shared[2048];
+ __shared__ float pad_temp_shared[4032];
+ __shared__ float kernel_shared[3072];
conv2d_nchw[0] = 0.000000e+00f;
- conv2d_nchw[2] = 0.000000e+00f;
- conv2d_nchw[4] = 0.000000e+00f;
- conv2d_nchw[6] = 0.000000e+00f;
- conv2d_nchw[8] = 0.000000e+00f;
- conv2d_nchw[10] = 0.000000e+00f;
- conv2d_nchw[12] = 0.000000e+00f;
- conv2d_nchw[1] = 0.000000e+00f;
- conv2d_nchw[3] = 0.000000e+00f;
- conv2d_nchw[5] = 0.000000e+00f;
conv2d_nchw[7] = 0.000000e+00f;
+ conv2d_nchw[1] = 0.000000e+00f;
+ conv2d_nchw[8] = 0.000000e+00f;
+ conv2d_nchw[2] = 0.000000e+00f;
conv2d_nchw[9] = 0.000000e+00f;
+ conv2d_nchw[3] = 0.000000e+00f;
+ conv2d_nchw[10] = 0.000000e+00f;
+ conv2d_nchw[4] = 0.000000e+00f;
conv2d_nchw[11] = 0.000000e+00f;
+ conv2d_nchw[5] = 0.000000e+00f;
+ conv2d_nchw[12] = 0.000000e+00f;
+ conv2d_nchw[6] = 0.000000e+00f;
conv2d_nchw[13] = 0.000000e+00f;
- for (int rc_outer_outer = 0; rc_outer_outer < 4; ++rc_outer_outer) {
- for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
- for (int rx_outer_outer = 0; rx_outer_outer < 3; ++rx_outer_outer) {
- __syncthreads();
- pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) - 8)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 48)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 104)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 168)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 160)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 216)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 280)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 272)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 328)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 392)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 384)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 448)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 440)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 504)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 496)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 552)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 616)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 608)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 664)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 728)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 720)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 776)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 840)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 832)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 888)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 952)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 944)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1008)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1000)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1064)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1056)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1112)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1176)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1168)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1232)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1224)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1288)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1280)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1344)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1336)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1400)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1392)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1456)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1448)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1512)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1504)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1568)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1560)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1624)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1616)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1680)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1672)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1736)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1728)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1792)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1784)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1848)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1840)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1904)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1896)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 1960)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 1952)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2016)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2008)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2072)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2064)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2128)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2120)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2184)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2176)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2240)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2232)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2296)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2288)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2352)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2344)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2408)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2400)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2464)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2456)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2520)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2512)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2576)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2568)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2632)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2624)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2688)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2680)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2744)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2736)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2800)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2792)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2856)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2848)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2912)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2904)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 2968)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 2960)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3024)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3016)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3080)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3072)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3136)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3128)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3192)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3184)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3248)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3240)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3304)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3296)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3360)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3352)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3416)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3408)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3472)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3464)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3528)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3520)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3584)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3576)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3640)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3632)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3696)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3688)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3752)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3744)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3808)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3800)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3864)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3856)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3920)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3912)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 3976)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 3968)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4032)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4024)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4088)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4080)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4144)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4136)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4200)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4192)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4256)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4248)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4312)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4304)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4368)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4360)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4424)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4416)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4480)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4472)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4536)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4528)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4592)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4584)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4648)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4640)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4704)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4696)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4760)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4752)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4816)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4808)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4872)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4864)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4928)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4920)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 4984)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 4976)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5040)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5032)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5096)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5088)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5152)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5144)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5208)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5200)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5264)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5256)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5320)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5312)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5376)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5368)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5432)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5424)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5488)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5480)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5544)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5536)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5600)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5592)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5656)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5648)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5712)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5704)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5768)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5760)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5824)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5816)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5880)] = (((((1 <= (((((int)threadIdx.x) % 49) / 7) + ry_outer_outer)) && ((((((int)threadIdx.x) % 49) / 7) + ry_outer_outer) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5872)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5936)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 1) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5928)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 5992)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 2) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 5984)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6048)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 3) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6040)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6104)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 4) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6096)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6160)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 5) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6152)] : 0.000000e+00f);
- pad_temp_shared[(((int)threadIdx.x) + 6216)] = (((((1 <= (ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7))) && ((ry_outer_outer + (((((int)threadIdx.x) / 7) + 6) % 7)) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 6272) + (ry_outer_outer * 7)) + ((int)threadIdx.x)) + rx_outer_outer) + 6208)] : 0.000000e+00f);
- kernel_shared[((int)threadIdx.x)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 56)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 112)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 112) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 168) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 224) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 96) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 280)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 280) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 216)];
- kernel_shared[(((int)threadIdx.x) + 336)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 336) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 80) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 392)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 392) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 72)];
- kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 448) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 576)];
- kernel_shared[(((int)threadIdx.x) + 504)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 504) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 120) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 560) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 432)];
- kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 616) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 104) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 672)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 672) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 288)];
- kernel_shared[(((int)threadIdx.x) + 728)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 728) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 88) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 784)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 784) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 144)];
- kernel_shared[(((int)threadIdx.x) + 840)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 840) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 648)];
- kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 32256)];
- kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 952) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1008) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1064) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1120) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 96) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1176) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 216)];
- kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1232) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 80) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1288) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 72)];
- kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1344) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 576)];
- kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1400) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 120) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1456) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 432)];
- kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1512) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 104) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1568) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 288)];
- kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1624) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 88) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1680) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 144)];
- kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1736) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 648)];
- kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 64512)];
- kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1848) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 504)];
- kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1904) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((((int)threadIdx.x) + 112) & 127) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer)];
- kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1960) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 360)];
- if (((int)threadIdx.x) < 32) {
- kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[(((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2016) >> 7) * 4608)) + (rc_outer_outer * 1152)) + (((int)threadIdx.x) * 9)) + (ry_outer_outer * 3)) + rx_outer_outer) + 864)];
- }
- __syncthreads();
- for (int rc_outer_inner = 0; rc_outer_inner < 4; ++rc_outer_inner) {
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[(((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32))]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 128)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 49)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 50)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 51)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 52)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 53)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 54)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 55)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 1)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 49)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 50)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 51)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 52)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 53)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 54)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 55)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 129)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 98)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 99)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 100)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 101)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 102)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 103)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 104)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 2)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 98)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 99)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 100)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 101)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 102)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 103)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 104)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 130)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 147)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 148)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 149)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 150)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 151)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 152)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 153)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 3)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 147)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 148)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 149)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 150)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 151)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 152)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 153)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 131)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 196)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 197)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 198)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 199)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 200)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 201)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 202)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 4)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 196)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 197)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 198)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 199)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 200)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 201)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 202)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 132)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 245)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 246)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 247)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 248)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 249)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 250)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 251)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 5)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 245)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 246)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 247)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 248)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 249)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 250)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 251)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 133)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 294)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 295)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 296)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 297)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 298)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 299)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 300)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 6)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 294)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 295)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 296)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 297)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 298)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 299)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 300)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 134)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 343)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 344)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 345)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 346)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 347)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 348)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 349)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 7)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 343)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 344)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 345)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 346)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 347)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 348)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 349)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 135)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 392)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 393)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 394)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 395)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 396)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 397)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 398)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 8)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 392)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 393)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 394)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 395)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 396)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 397)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 398)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 136)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 9)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 137)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 490)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 491)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 492)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 493)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 494)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 495)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 496)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 10)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 490)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 491)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 492)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 493)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 494)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 495)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 496)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 138)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 539)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 540)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 541)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 542)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 543)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 544)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 545)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 11)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 539)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 540)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 541)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 542)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 543)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 544)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 545)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 139)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 588)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 589)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 590)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 591)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 592)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 593)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 594)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 12)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 588)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 589)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 590)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 591)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 592)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 593)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 594)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 140)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 637)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 638)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 639)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 640)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 641)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 642)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 643)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 13)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 637)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 638)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 639)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 640)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 641)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 642)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 643)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 141)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 686)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 687)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 688)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 689)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 690)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 691)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 692)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 14)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 686)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 687)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 688)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 689)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 690)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 691)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 692)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 142)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 735)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 736)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 737)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 738)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 739)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 740)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 741)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 15)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 735)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 736)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 737)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 738)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 739)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 740)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 741)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 143)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 784)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 785)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 786)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 787)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 788)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 789)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 790)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 16)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 784)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 785)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 786)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 787)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 788)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 789)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 790)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 144)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 833)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 834)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 835)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 836)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 837)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 838)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 839)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 17)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 833)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 834)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 835)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 836)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 837)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 838)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 839)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 145)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 18)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 146)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 931)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 932)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 933)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 934)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 935)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 936)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 937)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 19)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 931)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 932)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 933)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 934)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 935)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 936)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 937)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 147)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 980)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 981)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 982)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 983)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 984)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 985)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 986)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 20)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 980)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 981)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 982)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 983)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 984)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 985)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 986)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 148)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1029)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1030)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1031)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1032)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1033)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1034)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1035)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 21)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1029)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1030)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1031)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1032)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1033)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1034)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1035)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 149)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1078)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1079)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1080)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1081)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1082)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1083)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1084)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 22)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1078)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1079)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1080)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1081)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1082)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1083)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1084)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 150)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1127)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1128)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1129)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1130)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1131)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1132)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1133)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 23)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1127)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1128)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1129)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1130)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1131)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1132)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1133)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 151)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1176)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1177)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1178)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1179)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1180)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1181)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1182)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 24)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1176)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1177)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1178)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1179)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1180)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1181)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1182)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 152)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1225)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1226)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1227)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1228)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1229)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1230)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1231)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 25)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1225)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1226)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1227)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1228)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1229)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1230)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1231)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 153)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1274)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1275)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1276)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1277)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1278)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1279)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1280)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 26)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1274)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1275)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1276)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1277)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1278)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1279)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1280)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 154)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1323)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1324)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1325)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1326)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1327)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1328)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1329)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 27)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1323)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1324)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1325)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1326)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1327)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1328)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1329)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 155)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1372)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1373)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1374)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1375)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1376)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1377)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1378)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 28)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1372)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1373)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1374)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1375)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1376)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1377)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1378)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 156)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1421)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1422)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1423)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1424)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1425)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1426)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1427)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 29)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1421)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1422)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1423)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1424)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1425)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1426)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1427)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 157)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1470)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1471)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1472)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1473)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1474)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1475)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1476)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 30)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1470)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1471)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1472)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1473)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1474)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1475)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1476)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 158)]));
- conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1519)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1520)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1521)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1522)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1523)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1524)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1525)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 31)]));
- conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1519)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1520)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1521)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1522)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1523)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1524)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
- conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[(((rc_outer_inner * 1568) + ((((int)threadIdx.x) % 7) * 7)) + 1525)] * kernel_shared[((((((int)threadIdx.x) / 7) * 256) + (rc_outer_inner * 32)) + 159)]));
+ for (int rc_outer_outer = 0; rc_outer_outer < 8; ++rc_outer_outer) {
+ for (int rx_outer_outer = 0; rx_outer_outer < 3; ++rx_outer_outer) {
+ __syncthreads();
+ pad_temp_shared[((int)threadIdx.x)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 56)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 56) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 112)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 112) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 168)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 168) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 224)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 224) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 280)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 280) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 336)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 336) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 392)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 392) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 448)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 448) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 504)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 384)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 560)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 560) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 616)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 616) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 672)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 672) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 728)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 728) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 784)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 784) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 840)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 840) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 896)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 896) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 952)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 952) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1008)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 776)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1064)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1064) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1120)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1120) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1176)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1176) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1232)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1232) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1288)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1288) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1344)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1344) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1400)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1400) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1456)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1456) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1512)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1168)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1568)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1568) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1624)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1624) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1680)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1680) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1736)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1736) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1792)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1792) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1848)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1848) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1904)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1904) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 1960)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 1960) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2016)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1560)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2072)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2072) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2128)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2128) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2184)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2184) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2240)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2240) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2296)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2296) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2352)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2352) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2408)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2408) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2464)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2464) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2520)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 1952)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2576)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2576) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2632)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2632) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2688)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2688) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2744)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2744) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2800)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2800) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2856)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2856) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2912)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2912) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 2968)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 2968) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3024)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 2344)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3080)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3080) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3136)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3136) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3192)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3192) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3248)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3248) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3304)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3304) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3360)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3360) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3416)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3416) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3472)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3472) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3528)] = ((((7 <= ((int)threadIdx.x)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((rc_outer_outer * 3136) + ((int)threadIdx.x)) + rx_outer_outer) + 2736)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3584)] = (((((1 <= (((((int)threadIdx.x) / 7) + 8) % 9)) && ((((((int)threadIdx.x) / 7) + 8) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3584) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 8) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3640)] = (((((1 <= (((((int)threadIdx.x) / 7) + 7) % 9)) && ((((((int)threadIdx.x) / 7) + 7) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3640) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 7) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3696)] = (((((1 <= (((((int)threadIdx.x) / 7) + 6) % 9)) && ((((((int)threadIdx.x) / 7) + 6) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3696) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 6) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3752)] = (((((1 <= (((((int)threadIdx.x) / 7) + 5) % 9)) && ((((((int)threadIdx.x) / 7) + 5) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3752) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 5) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3808)] = (((((1 <= (((((int)threadIdx.x) / 7) + 4) % 9)) && ((((((int)threadIdx.x) / 7) + 4) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3808) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 4) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3864)] = (((((1 <= (((((int)threadIdx.x) / 7) + 3) % 9)) && ((((((int)threadIdx.x) / 7) + 3) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3864) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 3) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3920)] = (((((1 <= (((((int)threadIdx.x) / 7) + 2) % 9)) && ((((((int)threadIdx.x) / 7) + 2) % 9) < 8)) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[((((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3920) / 63) * 49)) + ((((((int)threadIdx.x) / 7) + 2) % 9) * 7)) + rx_outer_outer) + (((int)threadIdx.x) % 7)) - 8)] : 0.000000e+00f);
+ pad_temp_shared[(((int)threadIdx.x) + 3976)] = ((((((int)threadIdx.x) < 49) && (1 <= (rx_outer_outer + (((int)threadIdx.x) % 7)))) && ((rx_outer_outer + (((int)threadIdx.x) % 7)) < 8)) ? data[(((((rc_outer_outer * 3136) + (((((int)threadIdx.x) + 3976) / 63) * 49)) + ((int)threadIdx.x)) + rx_outer_outer) - 1)] : 0.000000e+00f);
+ kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 56)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 112)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 168)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 168) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 224)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 224) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 280)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 280) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ kernel_shared[(((int)threadIdx.x) + 336)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 336) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 144) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 392)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 392) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 24)];
+ kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 448) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 192)];
+ kernel_shared[(((int)threadIdx.x) + 504)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 504) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 360)];
+ kernel_shared[(((int)threadIdx.x) + 560)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 560) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 176) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 616)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 616) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 120)];
+ kernel_shared[(((int)threadIdx.x) + 672)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 672) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 288)];
+ kernel_shared[(((int)threadIdx.x) + 728)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 728) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 152) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 784)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 784) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 840)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 840) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 216)];
+ kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 896) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 384)];
+ kernel_shared[(((int)threadIdx.x) + 952)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 952) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 184) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1008)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1008) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 144)];
+ kernel_shared[(((int)threadIdx.x) + 1064)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1064) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 312)];
+ kernel_shared[(((int)threadIdx.x) + 1120)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1120) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 160) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1176)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1176) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 72)];
+ kernel_shared[(((int)threadIdx.x) + 1232)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1232) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 240)];
+ kernel_shared[(((int)threadIdx.x) + 1288)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1288) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 408)];
+ kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 32256)];
+ kernel_shared[(((int)threadIdx.x) + 1400)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1400) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 1456)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1456) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 1512)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1512) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1568)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1568) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 1624)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1624) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ kernel_shared[(((int)threadIdx.x) + 1680)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1680) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 144) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1736)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1736) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 24)];
+ kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1792) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 192)];
+ kernel_shared[(((int)threadIdx.x) + 1848)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1848) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 360)];
+ kernel_shared[(((int)threadIdx.x) + 1904)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1904) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 176) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 1960)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 1960) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 120)];
+ kernel_shared[(((int)threadIdx.x) + 2016)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2016) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 288)];
+ kernel_shared[(((int)threadIdx.x) + 2072)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2072) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 152) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2128)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2128) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 48)];
+ kernel_shared[(((int)threadIdx.x) + 2184)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2184) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 216)];
+ kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2240) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 384)];
+ kernel_shared[(((int)threadIdx.x) + 2296)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2296) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 184) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2352)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2352) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 144)];
+ kernel_shared[(((int)threadIdx.x) + 2408)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2408) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 312)];
+ kernel_shared[(((int)threadIdx.x) + 2464)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2464) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 160) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2520)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2520) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 72)];
+ kernel_shared[(((int)threadIdx.x) + 2576)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2576) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 240)];
+ kernel_shared[(((int)threadIdx.x) + 2632)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2632) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 408)];
+ kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[(((((((int)blockIdx.x) * 73728) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 64512)];
+ kernel_shared[(((int)threadIdx.x) + 2744)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2744) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 168)];
+ kernel_shared[(((int)threadIdx.x) + 2800)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2800) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 336)];
+ kernel_shared[(((int)threadIdx.x) + 2856)] = kernel[(((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2856) / 192) * 4608)) + (rc_outer_outer * 576)) + (((((int)threadIdx.x) + 168) % 192) * 3)) + rx_outer_outer)];
+ kernel_shared[(((int)threadIdx.x) + 2912)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2912) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 96)];
+ kernel_shared[(((int)threadIdx.x) + 2968)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 2968) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 264)];
+ if (((int)threadIdx.x) < 48) {
+ kernel_shared[(((int)threadIdx.x) + 3024)] = kernel[((((((((int)blockIdx.x) * 73728) + (((((int)threadIdx.x) + 3024) / 192) * 4608)) + (rc_outer_outer * 576)) + (((int)threadIdx.x) * 3)) + rx_outer_outer) + 432)];
+ }
+ __syncthreads();
+ for (int rc_outer_inner = 0; rc_outer_inner < 4; ++rc_outer_inner) {
+ for (int ry_outer_inner = 0; ry_outer_inner < 3; ++ry_outer_inner) {
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[(((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[(((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7))] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 63)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 63)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 126)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 126)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 189)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 189)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 252)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 252)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 315)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 315)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 378)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 378)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 441)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 504)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 504)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 567)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 567)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 630)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 630)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 693)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 693)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 756)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 756)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 819)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 819)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 882)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 945)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 945)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 64)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 64)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 127)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 127)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 190)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 190)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 253)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 253)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 316)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 316)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 379)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 379)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 442)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 505)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 505)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 568)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 568)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 631)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 631)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 694)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 694)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 757)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 757)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 820)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 820)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 883)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 946)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 946)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 65)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 65)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 128)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 128)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 191)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 191)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 254)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 254)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 317)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 317)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 380)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 380)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 443)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 506)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 506)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 569)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 569)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 632)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 632)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 695)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 695)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 758)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 758)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 821)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 821)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 884)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 947)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 947)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 66)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 66)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 129)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 129)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 192)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 192)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 255)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 255)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 318)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 318)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 381)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 381)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 444)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 507)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 507)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 570)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 570)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 633)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 633)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 696)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 696)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 759)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 759)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 822)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 822)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 885)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 948)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 948)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 67)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 67)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 130)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 130)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 193)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 193)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 256)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 256)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 319)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 319)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 382)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 382)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 445)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 508)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 508)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 571)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 571)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 634)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 634)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 697)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 697)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 760)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 760)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 823)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 823)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 886)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 949)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 949)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 68)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 68)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 131)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 131)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 194)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 194)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 257)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 257)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 320)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 320)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 383)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 383)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 446)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 509)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 509)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 572)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 572)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 635)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 635)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 698)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 698)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 761)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 761)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 824)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 824)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 887)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 950)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 950)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1536)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 69)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 3)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 69)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1539)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 132)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 6)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 132)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1542)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 195)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 9)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 195)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1545)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 258)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 12)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 258)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1548)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 321)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 15)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 321)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1551)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 384)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 18)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 384)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1554)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 21)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 447)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1557)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 510)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 24)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 510)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1560)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 573)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 27)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 573)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1563)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 636)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 30)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 636)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1566)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 699)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 33)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 699)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1569)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 762)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 36)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 762)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1572)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 825)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 39)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 825)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1575)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 42)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 888)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1578)]));
+ conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 951)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 45)]));
+ conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[((((rc_outer_inner * 1008) + (ry_outer_inner * 7)) + ((((int)threadIdx.x) % 7) * 7)) + 951)] * kernel_shared[(((((((int)threadIdx.x) / 7) * 192) + (rc_outer_inner * 48)) + ry_outer_inner) + 1581)]));
}
}
}
}
- for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
- compute[((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7))] = max((conv2d_nchw[i1_inner] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 1)] = max((conv2d_nchw[(i1_inner + 2)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 2)] = max((conv2d_nchw[(i1_inner + 4)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 3)] = max((conv2d_nchw[(i1_inner + 6)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 4)] = max((conv2d_nchw[(i1_inner + 8)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 5)] = max((conv2d_nchw[(i1_inner + 10)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
- compute[(((((((int)blockIdx.x) * 784) + ((((int)threadIdx.x) / 7) * 98)) + (i1_inner * 49)) + ((((int)threadIdx.x) % 7) * 7)) + 6)] = max((conv2d_nchw[(i1_inner + 12)] + bias[(((((int)blockIdx.x) * 16) + ((((int)threadIdx.x) / 7) * 2)) + i1_inner)]), 0.000000e+00f);
+ for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
+ compute[(((((int)blockIdx.x) * 784) + (((int)threadIdx.x) * 7)) + i3_inner)] = max((conv2d_nchw[i3_inner] + bias[((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 7))]), 0.000000e+00f);
+ compute[((((((int)blockIdx.x) * 784) + (((int)threadIdx.x) * 7)) + i3_inner) + 392)] = max((conv2d_nchw[(i3_inner + 7)] + bias[(((((int)blockIdx.x) * 16) + (((int)threadIdx.x) / 7)) + 8)]), 0.000000e+00f);
}
}
</pre></div>
@@ -2109,7 +1585,7 @@ 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> ( 6 minutes 38.700 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 6 minutes 27.997 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 8b9555a2e4..8a92c1fa38 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -921,7 +921,7 @@ 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)
- 8.1015 8.1062 8.1119 8.0863 0.0110
+ 8.0760 8.0819 8.0826 8.0636 0.0088
</pre></div>
</div>
</div>
@@ -943,7 +943,7 @@ 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 12.714 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 12.916 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 9700a61d64..f429f9ae3d 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_x86.html
@@ -940,7 +940,7 @@ 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)
- 752.6929 752.2915 754.9909 750.7964 1.7358
+ 752.5467 752.2150 753.9007 751.5244 0.9981
</pre></div>
</div>
</div>
@@ -962,7 +962,7 @@ 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 44.702 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 46.055 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 cbc08fc35f..c3927d017f 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_sparse_x86.html
@@ -637,27 +637,27 @@ 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": T.bool(True), "global_symbol": "main", "tir.noalias": T.bool(True)})
- for i0_outer in T.parallel(4):
- compute_1 = T.allocate([1024], "float32", "global")
- for i1_outer in range(16):
- compute_2 = T.Buffer((1024,), data=compute_1)
- for i_outer_inner, nb_j_inner in T.grid(2, 2):
- for i_inner_init, j_init in T.grid(16, 16):
- compute_2[i_outer_inner * 512 + i_inner_init * 32 + nb_j_inner * 16 + j_init] = T.float32(0)
- for elem_idx, i_inner, j in T.grid(T.Let(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1], where={cse_var_1: i1_outer * 2 + nb_j_inner}), 16, 16):
- cse_var_1 = T.int32()
- placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
- cse_var_3: T.int32 = i1_outer * 2 + nb_j_inner
- cse_var_2: T.int32 = i_outer_inner * 512 + i_inner * 32 + nb_j_inner * 16 + j
+ for i0_outer_i1_outer_fused in T.parallel(512):
+ compute_1 = T.allocate([128], "float32", "global")
+ compute_2 = T.Buffer((128,), data=compute_1)
+ for i_outer_inner in range(2):
+ for i_inner_init, j_init in T.grid(4, 16):
+ compute_2[i_outer_inner * 64 + i_inner_init * 16 + j_init] = T.float32(0)
+ for elem_idx, i_inner, j in T.grid(T.Let(placeholder_5[cse_var_1 + 1] - placeholder_5[cse_var_1], where={cse_var_1: i0_outer_i1_outer_fused % 32}), 4, 16):
+ cse_var_1 = T.int32()
+ placeholder_5 = T.Buffer((33,), "int32", data=placeholder_3.data)
+ cse_var_2: T.int32 = i0_outer_i1_outer_fused % 32
+ if T.likely(elem_idx < placeholder_5[cse_var_2 + 1] - placeholder_5[cse_var_2]):
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 * 8192 + i_outer_inner * 4096 + 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 * 16384 + i0_inner * 512 + i1_outer * 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))
+ cse_var_3: T.int32 = i_outer_inner * 64 + i_inner * 16 + j
+ compute_2[cse_var_3] = compute_2[cse_var_3] + placeholder_6[placeholder_5[cse_var_2] * 16 + elem_idx * 16 + j] * T.max(placeholder_7[i0_outer_i1_outer_fused // 32 * 2048 + i_outer_inner * 1024 + i_inner * 256 + placeholder_8[placeholder_5[cse_var_2] + elem_idx]], T.float32(0))
+ for i0_inner in range(8):
+ cse_var_4: T.int32 = i0_outer_i1_outer_fused // 32 * 4096 + i0_inner * 512 + i0_outer_i1_outer_fused % 32 * 16
+ 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 + 16] = T.max(compute_2[i0_inner * 16:i0_inner * 16 + 16] + placeholder_5[cse_var_4:cse_var_4 + 16], T.Broadcast(T.float32(0), 16))
</pre></div>
</div>
</div>
@@ -691,7 +691,7 @@ 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: 1.521 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 1.336 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 4ae225e247..7f9447e3ea 100644
--- a/docs/how_to/tune_with_autotvm/sg_execution_times.html
+++ b/docs/how_to/tune_with_autotvm/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>00:55.789</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
+<p><strong>00:37.052</strong> total execution time for <strong>how_to_tune_with_autotvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -354,7 +354,7 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="tune_conv2d_cuda.html#sphx-glr-how-to-tune-with-autotvm-tune-conv2d-cuda-py"><span class="std std-ref">Tuning High Performance Convolution on NVIDIA GPUs</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_cuda.py</span></code>)</p></td>
-<td><p>00:55.753</p></td>
+<td><p>00:37.017</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tune_relay_x86.html#sphx-glr-how-to-tune-with-autotvm-tune-relay-x86-py"><span class="std std-ref">Auto-tuning a Convolutional Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_relay_x86.py</span></code>)</p></td>
diff --git a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
index d353d8d3a5..9a815b0373 100644
--- a/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
+++ b/docs/how_to/tune_with_autotvm/tune_conv2d_cuda.html
@@ -695,7 +695,7 @@ 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, 128, 1, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5738367
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 16, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2605244
No: 2 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -818,7 +818,7 @@ 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, 4, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1151832
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 64, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4000086
No: 3 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -941,9 +941,8 @@ 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, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4212501
-No: 4 GFLOPS: 1.02/1.02 result: MeasureResult(costs=(0.22650778175,), error_no=MeasureErrorNo.NO_ERROR, all_cost=5.790705442428589, timestamp=1682987906.4804852) [('tile_f', [-1, 8, 1, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3360248
-No: 5 GFLOPS: 0.00/1.02 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,533453
+No: 4 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1065,8 +1064,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8662698
-No: 6 GFLOPS: 0.00/1.02 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9991495
+No: 5 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1188,11 +1187,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8824596
-No: 7 GFLOPS: 30.20/30.20 result: MeasureResult(costs=(0.007664544214285715,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.030125617980957, timestamp=1682987910.191251) [('tile_f', [-1, 1, 2, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4294570
-No: 8 GFLOPS: 161.19/161.19 result: MeasureResult(costs=(0.0014362217428571429,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.2537076473236084, timestamp=1682987911.046372) [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5295007
-No: 9 GFLOPS: 28.82/161.19 result: MeasureResult(costs=(0.0080316095,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.9334707260131836, timestamp=1682987922.23613) [('tile_f', [-1, 1, 1, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6271065
-No: 10 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 8, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6152202
+No: 6 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1314,26 +1310,8 @@ 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, 16, 8, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5508265
-No: 11 GFLOPS: 0.00/161.19 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, 32, 2, 1]), ('tile_y', [-1, 7, 1, 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', 1)],None,9297435
-No: 12 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 1, 1, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,188704
+No: 7 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1455,8 +1433,8 @@ 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, 128, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7835177
-No: 13 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 32, 1, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8251241
+No: 8 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1578,8 +1556,8 @@ 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, 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', 1500), ('unroll_explicit', 0)],None,4548917
-No: 14 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 128, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3263751
+No: 9 GFLOPS: 0.00/0.00 result: Traceback (most recent call last):
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1701,8 +1679,9 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5217649
-No: 15 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9042267
+No: 10 GFLOPS: 40.99/40.99 result: MeasureResult(costs=(0.005647334888888888,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.6838254928588867, timestamp=1683002350.2191744) [('tile_f', [-1, 1, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2784230
+No: 11 GFLOPS: 0.00/40.99 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
@@ -1824,9 +1803,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 1, 8]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7311838
-No: 16 GFLOPS: 7.14/161.19 result: MeasureResult(costs=(0.03242240125,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.5403568744659424, timestamp=1682987927.5691445) [('tile_f', [-1, 4, 1, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 64]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7716502
-No: 17 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10048353
+No: 12 GFLOPS: 0.00/40.99 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
@@ -1948,8 +1926,776 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 2, 2, 128]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4791154
-No: 18 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 4, 2, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1152285
+No: 13 GFLOPS: 0.00/40.99 result: Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 742, in __call__
+ yield remote, remote.load_module(os.path.split(build_result.filename)[1])
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
+ costs = time_f(*args).results
+ File "/workspace/python/tvm/runtime/module.py", line 399, in evaluator
+ blob = feval(*args)
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 262, in tvm._ffi._cy3.core.FuncCall
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 251, in tvm._ffi._cy3.core.FuncCall3
+ File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
+tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 4: TVMFuncCall
+ at ../src/runtime/c_runtime_api.cc:477
+ 3: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 2: tvm::runtime::RPCWrappedFunc::operator()(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../src/runtime/rpc/rpc_module.cc:129
+ 1: tvm::runtime::RPCClientSession::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)> const&)
+ at ../src/runtime/rpc/rpc_endpoint.cc:1012
+ 0: tvm::runtime::RPCEndpoint::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)>)
+ at ../src/runtime/rpc/rpc_endpoint.cc:804
+ File "../src/runtime/rpc/rpc_endpoint.cc", line 804
+TVMError:
+---------------------------------------------------------------
+An error occurred during the execution of TVM.
+For more information, please see: https://tvm.apache.org/docs/errors.html
+---------------------------------------------------------------
+ Check failed: (code == RPCCode::kReturn) is false: code=kShutdown
+
+During handling of the above exception, another exception occurred:
+
+Traceback (most recent call last):
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
+ costs = time_f(*args).results
+ File "/usr/lib/python3.7/contextlib.py", line 130, in __exit__
+ self.gen.throw(type, value, traceback)
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 746, in __call__
+ remote.remove(build_result.filename)
+ File "/workspace/python/tvm/rpc/client.py", line 144, in remove
+ self._remote_funcs["remove"] = self.get_function("tvm.rpc.server.remove")
+ File "/workspace/python/tvm/rpc/client.py", line 72, in get_function
+ return self._sess.get_function(name)
+ File "/workspace/python/tvm/runtime/module.py", line 179, in get_function
+ self.handle, c_str(name), ctypes.c_int(query_imports), ctypes.byref(ret_handle)
+ File "/workspace/python/tvm/_ffi/base.py", line 348, in check_call
+ raise get_last_ffi_error()
+tvm._ffi.base.TVMError: Traceback (most recent call last):
+ 52: 0xffffffffffffffff
+ 51: _start
+ 50: __libc_start_main
+ 49: _Py_UnixMain
+ 48: 0x0000000000650da0
+ 47: 0x0000000000650afa
+ 46: _PyFunction_FastCallDict
+ 45: _PyEval_EvalCodeWithName
+ 44: _PyEval_EvalFrameDefault
+ 43: _PyFunction_FastCallKeywords
+ 42: _PyEval_EvalCodeWithName
+ 41: _PyEval_EvalFrameDefault
+ 40: _PyMethodDef_RawFastCallKeywords
+ 39: 0x0000000000546369
+ 38: _PyEval_EvalCodeWithName
+ 37: _PyEval_EvalFrameDefault
+ 36: _PyFunction_FastCallKeywords
+ 35: _PyEval_EvalCodeWithName
+ 34: _PyEval_EvalFrameDefault
+ 33: _PyFunction_FastCallDict
+ 32: _PyEval_EvalCodeWithName
+ 31: _PyEval_EvalFrameDefault
+ 30: _PyObject_FastCallDict
+ 29: 0x00000000004c06e1
+ 28: _PyFunction_FastCallDict
+ 27: _PyEval_EvalFrameDefault
+ 26: _PyMethodDescr_FastCallKeywords
+ 25: 0x00000000005dcb58
+ 24: 0x00000000005dc83f
+ 23: 0x00000000004ba127
+ 22: _PyEval_EvalFrameDefault
+ 21: _PyFunction_FastCallKeywords
+ 20: _PyEval_EvalFrameDefault
+ 19: _PyFunction_FastCallKeywords
+ 18: _PyEval_EvalFrameDefault
+ 17: _PyFunction_FastCallKeywords
+ 16: _PyEval_EvalCodeWithName
+ 15: _PyEval_EvalFrameDefault
+ 14: 0x0000000000537c30
+ 13: _PyObject_FastCallKeywords
+ 12: 0x00007fd78302dfa2
+ 11: _ctypes_callproc
+ 10: ffi_call
+ 9: ffi_call_unix64
+ 8: TVMModGetFunction
+ at ../src/runtime/c_runtime_api.cc:408
+ 7: tvm::runtime::ModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, bool)
+ at ../src/runtime/module.cc:66
+ 6: tvm::runtime::RPCModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, tvm::runtime::ObjectPtr<tvm::runtime::Object> const&)
+ at ../src/runtime/rpc/rpc_module.cc:187
+ 5: tvm::runtime::RPCClientSession::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
+ at ../src/runtime/rpc/rpc_endpoint.cc:1007
+ 4: tvm::runtime::TVMRetValue tvm::runtime::RPCEndpoint::SysCallRemote<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(tvm::runtime::RPCCode, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
+ at ../src/runtime/rpc/rpc_endpoint.h:223
+ 3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(int&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) const
+ at ../include/tvm/runtime/packed_func.h:1621
+ 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/rpc/rpc_endpoint.cc:684
+ File "../src/runtime/rpc/rpc_endpoint.cc", line 684
+TVMError:
+---------------------------------------------------------------
+An error occurred during the execution of TVM.
+For more information, please see: https://tvm.apache.org/docs/errors.html
+---------------------------------------------------------------
+ Check failed: (code == RPCCode::kReturn) is false: code=1
+
+Traceback (most recent call last):
+ 52: 0xffffffffffffffff
+ 51: _start
+ 50: __libc_start_main
+ 49: _Py_UnixMain
+ 48: 0x0000000000650da0
+ 47: 0x0000000000650afa
+ 46: _PyFunction_FastCallDict
+ 45: _PyEval_EvalCodeWithName
+ 44: _PyEval_EvalFrameDefault
+ 43: _PyFunction_FastCallKeywords
+ 42: _PyEval_EvalCodeWithName
+ 41: _PyEval_EvalFrameDefault
+ 40: _PyMethodDef_RawFastCallKeywords
+ 39: 0x0000000000546369
+ 38: _PyEval_EvalCodeWithName
+ 37: _PyEval_EvalFrameDefault
+ 36: _PyFunction_FastCallKeywords
+ 35: _PyEval_EvalCodeWithName
+ 34: _PyEval_EvalFrameDefault
+ 33: _PyFunction_FastCallDict
+ 32: _PyEval_EvalCodeWithName
+ 31: _PyEval_EvalFrameDefault
+ 30: _PyObject_FastCallDict
+ 29: 0x00000000004c06e1
+ 28: _PyFunction_FastCallDict
+ 27: _PyEval_EvalFrameDefault
+ 26: _PyMethodDescr_FastCallKeywords
+ 25: 0x00000000005dcb58
+ 24: 0x00000000005dc83f
+ 23: 0x00000000004ba127
+ 22: _PyEval_EvalFrameDefault
+ 21: _PyFunction_FastCallKeywords
+ 20: _PyEval_EvalFrameDefault
+ 19: _PyFunction_FastCall [('tile_f', [-1, 32, 1, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,4109301
+No: 14 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 8, 32]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,657998
+No: 15 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 16, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 256]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,380668
+No: 16 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 128, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10092522
+No: 17 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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, 4, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2766793
+No: 18 GFLOPS: 0.00/40.99 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=target, 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 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:1734
+ 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:1674
+ 19: run<>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 18: run<tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 17: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 16: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 15: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1634
+ 14: run<tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_, tvm::runtime::TVMMovableArgValueWithContext_>
+ at ../include/tvm/runtime/packed_func.h:1649
+ 13: operator()
+ at ../src/driver/driver_api.cc:401
+ 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:387
+ 11: tvm::LowerWithPassList(tvm::IRModule, tvm::runtime::Array<tvm::transform::Pass, void>)
+ at ../src/driver/driver_api.cc:282
+ 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:101
+ 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:1753
+ 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:1697
+ 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:1621
+ 2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
+ at ../include/tvm/runtime/packed_func.h:1217
+ 1: Call
+ at ../include/tvm/runtime/packed_func.h:1213
+ 0: operator()
+ at ../src/runtime/c_runtime_api.cc:534
+ File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
+ File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
+ raise InstantiationError("Skipped because of invalid gpu kernel")
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 8, 4, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7565072
+No: 19 GFLOPS: 0.00/40.99 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
@@ -2071,9 +2817,8 @@ Traceback (most recent call last):
File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
raise InstantiationError("Skipped because of invalid gpu kernel")
-tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 8, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1635073
-No: 19 GFLOPS: 16.89/161.19 result: MeasureResult(costs=(0.013710141181818182,), error_no=MeasureErrorNo.NO_ERROR, all_cost=8.34546685218811, timestamp=1682987936.1495934) [('tile_f', [-1, 2, 4, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5174693
-No: 20 GFLOPS: 0.00/161.19 result: Traceback (most recent call last):
+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, 2, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2618857
+No: 20 GFLOPS: 0.00/40.99 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
@@ -2195,7 +2940,7 @@ 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, 64, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 16]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1477150
+tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel [('tile_f', [-1, 64, 2, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,564976
</pre></div>
</div>
<p>Finally we can inspect the best config from log file, check correctness,
@@ -2234,9 +2979,9 @@ and measure running time.</p>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Finish loading 20 records
Best config:
-[('tile_f', [-1, 4, 64, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5295007
+[('tile_f', [-1, 1, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2784230
Finish loading 20 records
-Time cost of this operator: 0.002261
+Time cost of this operator: 0.006015
</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 54ee4a53b2..25014f3bc9 100644
--- a/docs/how_to/work_with_microtvm/micro_autotune.html
+++ b/docs/how_to/work_with_microtvm/micro_autotune.html
@@ -649,10 +649,10 @@ 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 316.2 98.733 (1, 2, 10, 10, 3) 2 1 [316.2]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.099 0.968 (1, 6, 10, 10) 1 1 [3.099]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.959 0.299 (1, 1, 10, 10, 3) 1 1 [0.959]
-Total_time - 320.257 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 315.6 98.743 (1, 2, 10, 10, 3) 2 1 [315.6]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 3.052 0.955 (1, 6, 10, 10) 1 1 [3.052]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.966 0.302 (1, 1, 10, 10, 3) 1 1 [0.966]
+Total_time - 319.618 - - - - -
</pre></div>
</div>
</div>
@@ -704,13 +704,13 @@ 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 138.4 98.128 (1, 6, 10, 10, 1) 2 1 [138.4]
-tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.8 1.276 (1, 6, 10, 10) 1 1 [1.8]
-tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.841 0.596 (1, 3, 10, 10, 1) 1 1 [0.841]
-Total_time - 141.041 - - - - -
+tvmgen_default_fused_nn_contrib_conv2d_NCHWc tvmgen_default_fused_nn_contrib_conv2d_NCHWc 101.3 97.383 (1, 6, 10, 10, 1) 2 1 [101.3]
+tvmgen_default_fused_layout_transform_1 tvmgen_default_fused_layout_transform_1 1.772 1.703 (1, 6, 10, 10) 1 1 [1.772]
+tvmgen_default_fused_layout_transform tvmgen_default_fused_layout_transform 0.951 0.914 (1, 1, 10, 10, 3) 1 1 [0.951]
+Total_time - 104.022 - - - - -
</pre></div>
</div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 25.585 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 24.486 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-work-with-microtvm-micro-autotune-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../../_downloads/9ccca8fd489a1486ac71b55a55c320c5/micro_autotune.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_autotune.py</span></code></a></p>
diff --git a/docs/how_to/work_with_microtvm/micro_pytorch.html b/docs/how_to/work_with_microtvm/micro_pytorch.html
index b2e83e65bc..5b60168bf1 100644
--- a/docs/how_to/work_with_microtvm/micro_pytorch.html
+++ b/docs/how_to/work_with_microtvm/micro_pytorch.html
@@ -460,7 +460,8 @@ download a cat image and preprocess it to use as the model input.</p>
Downloading: "https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
0%| | 0.00/3.42M [00:00<?, ?B/s]
-100%|##########| 3.42M/3.42M [00:00<00:00, 72.9MB/s]
+ 62%|######2 | 2.13M/3.42M [00:00<00:00, 22.4MB/s]
+100%|##########| 3.42M/3.42M [00:00<00:00, 33.8MB/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.
@@ -586,7 +587,7 @@ 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 20.376 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 20.458 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 b0215e1b42..c1cf6dcef9 100644
--- a/docs/how_to/work_with_microtvm/micro_train.html
+++ b/docs/how_to/work_with_microtvm/micro_train.html
@@ -528,7 +528,7 @@ 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/tmp293dho4g/images/random'
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>'/tmp/tmp0h6ga6xi/images/random'
</pre></div>
</div>
</div>
@@ -588,8 +588,8 @@ 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], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0]" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>/tmp/tmp293dho4g/images/target contains 8144 images
-/tmp/tmp293dho4g/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], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [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/tmp0h6ga6xi/images/target contains 8144 images
+/tmp/tmp0h6ga6xi/images/random contains 5000 images
</pre></div>
</div>
</div>
@@ -701,13 +701,13 @@ 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 - 41s - loss: 0.2148 - accuracy: 0.9258 - val_loss: 0.2106 - val_accuracy: 0.9339 - 41s/epoch - 124ms/step
+328/328 - 40s - loss: 0.2085 - accuracy: 0.9287 - val_loss: 0.1437 - val_accuracy: 0.9562 - 40s/epoch - 122ms/step
Epoch 2/3
-328/328 - 35s - loss: 0.0970 - accuracy: 0.9655 - val_loss: 0.1369 - val_accuracy: 0.9524 - 35s/epoch - 107ms/step
+328/328 - 35s - loss: 0.0874 - accuracy: 0.9671 - val_loss: 0.1198 - val_accuracy: 0.9539 - 35s/epoch - 108ms/step
Epoch 3/3
-328/328 - 35s - loss: 0.0602 - accuracy: 0.9768 - val_loss: 0.1027 - val_accuracy: 0.9622 - 35s/epoch - 106ms/step
+328/328 - 35s - loss: 0.0651 - accuracy: 0.9783 - val_loss: 0.1155 - val_accuracy: 0.9641 - 35s/epoch - 107ms/step
-<keras.callbacks.History object at 0x7fceb538fbd0>
+<keras.callbacks.History object at 0x7fa88e2ae510>
</pre></div>
</div>
</div>
@@ -971,7 +971,7 @@ 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 32.894 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 4 minutes 28.517 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 3819abcc0f..1235ac4282 100644
--- a/docs/how_to/work_with_microtvm/sg_execution_times.html
+++ b/docs/how_to/work_with_microtvm/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>07:45.319</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
+<p><strong>07:39.833</strong> total execution time for <strong>how_to_work_with_microtvm</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 82%" />
@@ -354,27 +354,27 @@
</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">5. 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>04:32.894</p></td>
+<td><p>04:28.517</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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">6. Model Tuning with microTVM</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_autotune.py</span></code>)</p></td>
-<td><p>01:25.585</p></td>
+<td><p>01:24.486</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><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">4. microTVM PyTorch Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_pytorch.py</span></code>)</p></td>
-<td><p>01:20.376</p></td>
+<td><p>01:20.458</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">3. microTVM Ahead-of-Time (AOT) Compilation</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_aot.py</span></code>)</p></td>
-<td><p>00:10.628</p></td>
+<td><p>00:10.629</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_custom_ide.html#sphx-glr-how-to-work-with-microtvm-micro-custom-ide-py"><span class="std std-ref">9. Bring microTVM to your own development environment</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_custom_ide.py</span></code>)</p></td>
-<td><p>00:08.325</p></td>
+<td><p>00:08.270</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><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">2. microTVM TFLite Tutorial</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_tflite.py</span></code>)</p></td>
-<td><p>00:07.512</p></td>
+<td><p>00:07.473</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="micro_ethosu.html#sphx-glr-how-to-work-with-microtvm-micro-ethosu-py"><span class="std std-ref">7. Running TVM on bare metal Arm(R) Cortex(R)-M55 CPU and Ethos(TM)-U55 NPU with CMSIS-NN</span></a> (<code class="docutils literal notranslate"><span class="pre">micro_ethosu.py</span></code>)</p></td>
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 90a8016019..a43b1d70d2 100644
--- a/docs/how_to/work_with_relay/sg_execution_times.html
+++ b/docs/how_to/work_with_relay/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>00:37.268</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
+<p><strong>00:37.443</strong> total execution time for <strong>how_to_work_with_relay</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 84%" />
@@ -354,15 +354,15 @@
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="using_pipeline_executor.html#sphx-glr-how-to-work-with-relay-using-pipeline-executor-py"><span class="std std-ref">Using Pipeline Executor in Relay</span></a> (<code class="docutils literal notranslate"><span class="pre">using_pipeline_executor.py</span></code>)</p></td>
-<td><p>00:32.528</p></td>
+<td><p>00:32.468</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>00:02.931</p></td>
+<td><p>00:02.941</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>00:01.802</p></td>
+<td><p>00:02.028</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 3a423d4560..45ad857e9b 100644
--- a/docs/how_to/work_with_schedules/intrin_math.html
+++ b/docs/how_to/work_with_schedules/intrin_math.html
@@ -554,7 +554,7 @@ 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 0x7fcafe5d4950>
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span><function my_cuda_math_rule at 0x7fa4d2daf200>
</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 5d3f0c6d8d..63dcb5f584 100644
--- a/docs/how_to/work_with_schedules/sg_execution_times.html
+++ b/docs/how_to/work_with_schedules/sg_execution_times.html
@@ -345,7 +345,7 @@
<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>00:09.334</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
+<p><strong>00:08.385</strong> total execution time for <strong>how_to_work_with_schedules</strong> files:</p>
<table class="docutils align-default">
<colgroup>
<col style="width: 83%" />
@@ -354,31 +354,31 @@
</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>00:06.570</p></td>
+<td><p>00:05.696</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>00:01.246</p></td>
+<td><p>00:01.190</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>00:00.624</p></td>
+<td><p>00:00.617</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>00:00.605</p></td>
+<td><p>00:00.603</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>00:00.136</p></td>
+<td><p>00:00.129</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tedd.html#sphx-glr-how-to-work-with-schedules-tedd-py"><span class="std std-ref">Use Tensor Expression Debug Display (TEDD) for Visualization</span></a> (<code class="docutils literal notranslate"><span class="pre">tedd.py</span></code>)</p></td>
-<td><p>00:00.066</p></td>
+<td><p>00:00.064</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="schedule_primitives.html#sphx-glr-how-to-work-with-schedules-schedule-primitives-py"><span class="std std-ref">Schedule Primitives in TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">schedule_primitives.py</span></code>)</p></td>
-<td><p>00:00.058</p></td>
+<td><p>00:00.056</p></td>
<td><p>0.0 MB</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="tuple_inputs.html#sphx-glr-how-to-work-with-schedules-tuple-inputs-py"><span class="std std-ref">Compute and Reduce with Tuple Inputs</span></a> (<code class="docutils literal notranslate"><span class="pre">tuple_inputs.py</span></code>)</p></td>
diff --git a/docs/reference/api/python/auto_scheduler.html b/docs/reference/api/python/auto_scheduler.html
index 3fb3e720a2..d8df26a356 100644
--- a/docs/reference/api/python/auto_scheduler.html
+++ b/docs/reference/api/python/auto_scheduler.html
@@ -1622,7 +1622,7 @@ 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>
@@ -1906,7 +1906,7 @@ 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 086905a559..05a137789e 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/9530a8de1/web/src/rpc_server.ts#L45">rpc_server.ts:45</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L45">rpc_server.ts:45</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/9530a8de1/web/src/rpc_server.ts#L45">rpc_server.ts:45</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L45">rpc_server.ts:45</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/9530a8de1/web/src/rpc_server.ts#L44">rpc_server.ts:44</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L44">rpc_server.ts:44</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/9530a8de1/web/src/rpc_server.ts#L65">rpc_server.ts:65</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L65">rpc_server.ts:65</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/9530a8de1/web/src/rpc_server.ts#L51">rpc_server.ts:51</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L51">rpc_server.ts:51</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/9530a8de1/web/src/rpc_server.ts#L59">rpc_server.ts:59</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/rpc_server.ts#L59">rpc_server.ts:59</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 c6a291f152..298e064568 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/9530a8de1/web/src/memory.ts#L223">memory.ts:223</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L208">memory.ts:208</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L312">memory.ts:312</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L284">memory.ts:284</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L388">memory.ts:388</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L376">memory.ts:376</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L267">memory.ts:267</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L243">memory.ts:243</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L321">memory.ts:321</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L252">memory.ts:252</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L359">memory.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L342">memory.ts:342</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L350">memory.ts:350</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L326">memory.ts:326</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L363">memory.ts:363</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L346">memory.ts:346</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/memory.ts#L334">memory.ts:334</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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 f41d8d6ddf..ab206c8723 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/9530a8de1/web/src/runtime.ts#L359">runtime.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L359">runtime.ts:359</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/9530a8de1/web/src/runtime.ts#L357">runtime.ts:357</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L357">runtime.ts:357</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/9530a8de1/web/src/runtime.ts#L355">runtime.ts:355</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L355">runtime.ts:355</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/9530a8de1/web/src/runtime.ts#L359">runtime.ts:359</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L359">runtime.ts:359</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/9530a8de1/web/src/runtime.ts#L376">runtime.ts:376</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L376">runtime.ts:376</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/9530a8de1/web/src/runtime.ts#L367">runtime.ts:367</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L367">runtime.ts:367</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 5f429372f9..7337df903c 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/9530a8de1/web/src/runtime.ts#L299">runtime.ts:299</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L299">runtime.ts:299</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/9530a8de1/web/src/runtime.ts#L297">runtime.ts:297</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L297">runtime.ts:297</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/9530a8de1/web/src/runtime.ts#L295">runtime.ts:295</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L295">runtime.ts:295</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/9530a8de1/web/src/runtime.ts#L320">runtime.ts:320</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L320">runtime.ts:320</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/9530a8de1/web/src/runtime.ts#L327">runtime.ts:327</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/runtime.ts#L327">runtime.ts:327</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 f37cbe80bd..0c70dd7d4d 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/9530a8de1/web/src/environment.ts#L86">environment.ts:86</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/environment.ts#L70">environment.ts:70</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/environment.ts#L69">environment.ts:69</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/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/9530a8de1/web/src/environment.ts#L78">environment.ts:78</a></li>
+ <li>Defined in <a href="https://github.com/apache/tvm/blob/c1e88b365/web/src/environment.ts#L78">environment.ts:78</a></li>
</ul>
... 2852 lines suppressed ...